You are on page 1of 302


Education I.II

Inge Schwank (Editor)

Forschungsinstitut für Mathematikdidaktik, Osnabrück 1999
ISBN: 3-925386-51-3
Die Deutsche Bibliothek - CIP-Einheitsaufnahme
Schwank, Inge: European Research in Mathematics Education I.II
- Proceedings of the First Conference of the European Society in
Mathematics Education Vol. II; Internet-Version
ISBN: 3-925386-51-3

Layout (including cover): By means of Corel Ventura ®
© Forschungsinstitut fuer Mathematikdidaktik, Osnabrueck

Please note:
Printed versions will be available:
Libri Books on Demand
Vol. I: ISBN: 3-925386-53-X
Vol. II: ISBN: 3-925386-54-8
Vol. III: ISBN: 3-925386-55-6

Vol. I: ISBN: 3-925386-50-5
Vol. II: ISBN: 3-925386-51-3
Vol. III: ISBN: 3-925386-52-1
European Research in
Mathematics Education I.II

Proceedings of the First Conference of the
European Society for Research in Mathematics Education, Vol. II

Inge Schwank (Editor)

Forschungsinstitut für Mathematikdidaktik, Osnabrück 1999
ISBN: 3-925386-51-3
European Research in Mathematics Education I.II 4



AS COGNITIVE PROCESSES . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Inge Schwank, Emanuila Gelfman, Elena Nardi

The Nature of Cognitive Structures
- Introduction and Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
General framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Individual Preferences - Predicative versus Functional Cognitive Structures 18
Cognitive Experiences - Information Coding. . . . . . . . . . . . . . . . . 19
Individual Styles of Cognition . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Cognition and Emotion / Motivation . . . . . . . . . . . . . . . . . . . . . . . 20
Cognition and Language . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Cognition and New Technologies . . . . . . . . . . . . . . . . . . . . . . . . . 21
Mathematical Thinking in Modern Conditions – Situated Cognition . . . . . . . 22
Benefit and Outcome of the Working Group . . . . . . . . . . . . . . . . . . . 22

IN ADVANCED MATHEMATICS LEARNING . . . . . . . . . . . . . . . . 24
Pier Luigi Ferrari

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
The Interpretation of Inequality Symbol . . . . . . . . . . . . . . . . . . . . . 27
A Divisibility Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Terms and Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
European Research in Mathematics Education I.II 5

The Role of Problem Presentation in Problem Solving. . . . . . . . . . . . 31
Some Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

INTELLECTUAL DEVELOPMENT . . . . . . . . . . . . . . . . . . . . . . . 38
Emanuila Gelfman, Marina Kholodnaya

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Ways of Information Coding as One of the Conditions of Growth
of Conceptual Competence . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Ways of Information Coding as One of Conditions of Individualisation
of Intellectual Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

Emanuila Gelfman, Raissa Lozinskaia, Galina Remezova

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

Marie Kubinova, Jarmila Novotna, Graham H. Littler

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Our Approach to Projects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
What Is a Project . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Work With Projects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
Work With Our Tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
European Research in Mathematics Education I.II 6

THREE-TERM COMPARISONS . . . . . . . . . . . . . . . . . . . . . . . . . 64
Pearla Nesher, Sara Hershkovitz, Jarmila Novotna

The Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Analysis of the Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Variables Involved in the Study . . . . . . . . . . . . . . . . . . . . . . . . . . 67
The Compared and the Reference . . . . . . . . . . . . . . . . . . . . . . 67
“More Than” or “Less Than” . . . . . . . . . . . . . . . . . . . . . . . . . 68
The Underlying Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
The Variables of the Study:. . . . . . . . . . . . . . . . . . . . . . . . . . 70
The Experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73

WITH TECHNOLOGIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Tatyana Oleinik

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
The Experiments and Method . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Some Remarks About Essence of Students’ Researches . . . . . . . . . . . 79
Some Preliminary Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83

Inge Schwank

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Predicative versus Functional Cognitive Structures . . . . . . . . . . . . . . . . 88
Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Fitting Figures in Matrices: QuaDiPF . . . . . . . . . . . . . . . . . . . . 89
EEG-Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Eye Movement - Reflections. . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
European Research in Mathematics Education I.II 7

Outlook. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

Rosetta Zan, Paola Poli

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104


coordinated by Paolo Boero

General Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
Summary of Discussions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
The Nature of Algebra Related to the Teaching of Algebra . . . . . . . . 107
Arithmetic and Algebra . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Tools of Research in Teaching and Learning of Algebra . . . . . . . . . . 109
Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Construction and Interpretation of Algebraic Expressions Related
to Other Domains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
General Comments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
European Research in Mathematics Education I.II 8

SYMBOLIC EXPRESSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Luciana Bazzini

The problem’s Position . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
The Teaching Experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Problem I: Analysis of the Students’ Answers . . . . . . . . . . . . . . . 115
Problem II: Analysis of the Students’ Answers . . . . . . . . . . . . . . . 116
Question I): Analysis of the Students’ Answers. . . . . . . . . . . . . . . 117
Question II): Analysis of the Students’ Answers . . . . . . . . . . . . . . 119
Final Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

FROM SENSE TO SYMBOL SENSE . . . . . . . . . . . . . . . . . . . . . . 123
Christer Bergsten

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Mathematical Forms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
Mathematics and Image Schemata . . . . . . . . . . . . . . . . . . . . . . . . 127
Mathematical Symbolism and Image Schemata . . . . . . . . . . . . . . . . . 129
Developmental Aspects of Understanding Forms . . . . . . . . . . . . . . . . 131
Symbol Sense . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

MATHEMATICAL ORGANIZATION . . . . . . . . . . . . . . . . . . . . . 135
Pilar Bolea, Marianna Bosch, Josep Gascón

Mathematics as the Study of Mathematical Works . . . . . . . . . . . . . . . 135
The Need of a Model of School Algebra . . . . . . . . . . . . . . . . . . . . . 137
The Algebraization of a Mathematical Work. . . . . . . . . . . . . . . . . . . 139
Characteristics of an Algebraized Mathematical Organization . . . . . . . . . 141
The Role of Algebraization in the Study of a Mathematical Work . . . . . . . 143
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
European Research in Mathematics Education I.II 9

Veronica Hoyos, Bernard Capponi, B. Geneves

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
Class Room Context . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Theoretical Frame of our Educational Experience . . . . . . . . . . . . . . . . 148
The Sources and the Techniques . . . . . . . . . . . . . . . . . . . . . . . . . 149
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154

Nicolina A. Malara, Rosa Iaderosa

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
A Few Considerations on Syntactic Difficulties in the Learning of Algebra . . 161
The Confusion Between Additive and Multiplicative Notation:
A Hypothesis of Inquiry for its Overcoming . . . . . . . . . . . . . . . . . . . 164
Considerations and Problematization of Some Arithmetical-Algebraic
Activities in Middle School . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170

A STUDY WITH 14 YEAR OLD PUPILS . . . . . . . . . . . . . . . . . . . 172
Maria Reggiani

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
Research Context . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
Research Problem. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
Research Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Summary of the Work Carried out in Class . . . . . . . . . . . . . . . . . . . 175
Verification Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
Analysis of the Proposed Problem . . . . . . . . . . . . . . . . . . . . . . . . 177
European Research in Mathematics Education I.II 10

Analysis of Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Analysis of Single Protocols . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
Final Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183

Sam Rososhek

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Weyl-Shafarevich Conception . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Basic Elements of Algebraic Knowledge . . . . . . . . . . . . . . . . . . . . 185
Algebra Propaedeutics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Mathematical Modelling Line . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Algebraic Structures Introduction . . . . . . . . . . . . . . . . . . . . . . . . 189
Computer Visualization of Algebraic Structures . . . . . . . . . . . . . . . . 192
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193

Éva Szeredi, Judit Török

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
A Priori Analysis of the Task. . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Some Possible Solutions. . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Reference Knowledge . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Processes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
Difficulties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
A Priori Expectations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
Analysis of Performances . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
How the Students Exploited or not Reference Knowledge . . . . . . . . . 200
Processes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
Comparison Between Hungarian and Italian Students . . . . . . . . . . . . . . 207
Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Possible Interpretations . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
European Research in Mathematics Education I.II 11

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209


Milan Hejny, Christine Shiu, Juan Diaz Godino, Herman Maier

Theories in and of Mathematics Education. . . . . . . . . . . . . . . . . . . . 213
What is a Theory and How Do Theories Arise?. . . . . . . . . . . . . . . 213
What Is the Role and Function of Theory in the Research Process? . . . . 213
Is It Feasible and Useful to Distinguish Different Kinds of Theory? . . . . 215
How Can Theories in Mathematics Education Be Evaluated? . . . . . . . 215
Conducting Empirical Research . . . . . . . . . . . . . . . . . . . . . . . . . 216
What Can Be Learned from Empirical Research? . . . . . . . . . . . . . 216
What Is the Relationship Between the Researcher and the Researched? . . 217
How Can Research Best Be Shared with Teachers? . . . . . . . . . . . . 217
Developing Methodologies . . . . . . . . . . . . . . . . . . . . . . . . . 217
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
Communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
Cooperation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Collaboration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
European Research in Mathematics Education I.II 12

EDUCATION: REFLECTIONS AND EXAMPLES . . . . . . . . . . . . . . 220
Giorgio T. Bagni

Bombelli’s Algebra (1572) and Imaginary Numbers. . . . . . . . . . . . . . . 220
Educational Problems: The Focus of Our Work . . . . . . . . . . . . . . . . . 224
The Group Concept from History to Mathematics Education . . . . . . . . . . 226
History of Mathematics and Epistemology of Learning . . . . . . . . . . . . . 229
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230

UNITS FOR DIDACTIC OF MATHEMATICS. . . . . . . . . . . . . . . . . 232
Juan Diaz Godino, Carmen Batanero

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
Meaning as the Content of Semiotic Functions . . . . . . . . . . . . . . . . . 233
Notational, Extensional and Intensional Meanings. . . . . . . . . . . . . . . . 234
Elementary and Systemic Meanings . . . . . . . . . . . . . . . . . . . . . . . 236
Personal and Institutional Meanings . . . . . . . . . . . . . . . . . . . . . . . 236
Semiotic Acts and Processes in the Study of Numbers . . . . . . . . . . . . . 237
Elementary Meanings: The First Encounter With Numbers . . . . . . . . 237
Systemic Meanings of Numbers. . . . . . . . . . . . . . . . . . . . . . . 240
Conclusions and Implications . . . . . . . . . . . . . . . . . . . . . . . . . . 242
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244

José Luis González Marí

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
Mathematics Education: A Field of Relationships vs. Related Fields . . . . . . 248
Insufficiency of the Interdisciplinary Approach and Specificity . . . . . . . . . 250
Didactical Analysis of a Mathematical Knowledge . . . . . . . . . . . . . . . 252
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
European Research in Mathematics Education I.II 13

EIGENPRODUCTIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Bruno D’Amore, Hermann Maier

Pupils’ Textual Eigenproductions . . . . . . . . . . . . . . . . . . . . . . . . 257
Characterisation of Textual Eigenproduction . . . . . . . . . . . . . . . . 257
Didactical Function of Textual Eigenproduction . . . . . . . . . . . . . . 258
Teachers Interpretation of Textual Eigenproductions . . . . . . . . . . . . . . 260
Design of the Research . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
Some Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273

André Rouchier

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
The Classical Components of the Theory of Didactical Situations . . . . . . . 276
Knowledge and Institutions, the Particular Case of Mathematics . . . . . . . . 278
New Objects, New Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . 280
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284

Nada Stehlíková

Introduction and Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
Research Projects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
A Student’s Solution to a Word Problem . . . . . . . . . . . . . . . . . . 286
Ability to Construct a Mental Representation of a Structure . . . . . . . . 286
Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
How to Create Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
Phenomena . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
Trace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
European Research in Mathematics Education I.II 14

Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
Inner Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294
Implementing Research Results . . . . . . . . . . . . . . . . . . . . . . . . . 295
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296



European Research in Mathematics Education I.II: Group 6 16

Inge Schwank1, Emanuila Gelfman2, Elena Nardi3
Institute for Cognitive Mathematics, Department of Mathematics & Computer Science,
University of Osnabrueck, D-49069 Osnabrueck
Department of Algebra & Geometry, Tomsk Pedagogical University,
Komsomolskii Prospect, 75, 634041Tomsk, Russia
University of Oxford, 15, Norham Gardens, OX2 6PY Oxford, U.K.

The work was prepared in such a way that - after the reviewing process - all accepted
papers were distributed to the prospected group members. In the spirit of CERME the
group leaders decided that during the sessions the accepted papers would not be orally
presented one by one. For the purpose of the stimulation of a goal-orientated, in depth
discussion six general themes had been identified and two members of the group were
asked to give a general introduction referring to the papers fitting each theme and to
current research developments. The themes and the introduction presenters are as

1. The Nature of Cognitive Structures - Introduction and Overview
Emanuila Gelfman & Inge Schwank
2. Individual Styles of Cognition
Sara Hershkovitz & Marina Kholodnaya
3. Cognition and Emotion/Motivation
Elena Nardi & Rosetta Zan
4. Cognition and Language
Pierre Luigi Ferrari & Pearla Nesher
5. Cognition and New Technologies
Tatiana Oleinik & Elmar Cohors-Fresenborg
6. Mathematical Thinking in Modern Conditions – Situated Cognition
Jarmila Novotna & Elena Nardi
European Research in Mathematics Education I.II: Group 6 17

u I
r Personal n
o Cognition
p Mathematical
- Knowledge t
h Representations / Tools - Problem Solving

- Thinking
s WW·WWsenso-motoric ! Other Fields
! (Ordinary) Language r
i Memory / Use & Change
! Beliefs
l Emotion / Motivation / Personality a

Ex Didactical Material
Teaching Strategies
ter !!! nal

Person 2 !!! Person n

Fig. 1: The Nature of Cognitive Structures - Personal Cognition
European Research in Mathematics Education I.II: Group 6 18

1. The Nature of Cognitive Structures
- Introduction and Overview

1.1 General framework

At the beginning of the group work Inge Schwank gave an introduction and developed a
general framework (Fig. 1) in which personal cognition could be considered under
different aspects.

One aspect was the role which different kinds of representation of concepts like
visual, verbal and sensor motoric play in the process of concept formation and how
these are correlated to mental tools. A second aspect dealt with the field of different
aspects of mathematical learning like the matter of mathematical knowledge, the
specific mental actions in problem solving and the nature of mathematical thinking. It
was worked out that these specific mathematical aspects had to be considered in their
correlation to other fields, especially cognition and the use of ordinary language. These
different aspects of one person’s cognition are controlled by the person’s beliefs. This
is the bridge between cognition and the role of emotion, motivation and the person’s
personality. Inge Schwank mentioned that these psychological reflections had to be
based on some outcomes of neurophysiology and, even more, on neurobiology. All
these aspects discuss cognition under the internal perspective of a person. Of course, in
studying mathematical teaching and learning processes this point of view has to be
broadened to an external perspective in which different aspects of mathematics,
didactical materials or teacher strategies are some examples which influence one
person’s cognitive processes. A broader perspective comes into consideration if one
studies interactions of persons, either among learners or between teachers and learners.

1.2 Individual Preferences
- Predicative versus Functional Cognitive Structures

In the centre of discussion there was Inge Schwank’s theory concerning the distinction
of predicative versus functional cognitive structures and the hypothesis that most
people have a more or less strong preference for one of these two cognitive structures in
European Research in Mathematics Education I.II: Group 6 19

which they model, understand and solve given external problems. Inge Schwank
pointed out that this distinction had its roots in different aspects of mathematics. As one
example she explained how famous mathematicians stress the aspects that
mathematical concepts have to be understood through the glasses of process orientation
and that concepts very often play the role of tools and not of statements. Relations to
some of the closest theoretical concepts were discussed, e.g. declarative - procedural
knowledge, procepts, APO-Schema (action®process®object - rule).

1.3 Cognitive Experiences - Information Coding

Emanuila Gelfman treated composition of some forms of students’ mental experience:
cognitive, metacognitive and intentional. Students’ cognitive experience is paid special
attention. In particular, Emanuila Gelfman gave the examples of how, using specially
constructed school-texts for 10-15 year old students, we may introduce different forms
of information coding: verbal, visual, sensual-sensory. Then she pointed out the role
which different ways of information coding plays in students’ intellectual
developments and dealt with the advantages and difficulties which arise from verbal
versus formal representation of mathematical concepts and problems.

2. Individual Styles of Cognition

This aspect of cognition was tackled under two complementary points of view. Sara
Hershkovitz started her presentation from examples of analysing pupils’ behaviour
when they are dealing with specific word problems.

In a more general talk Marina Kholodnaya explained - from a psychological point of
view - which different aspects of individual styles had to be considered: information
coding styles, information processing styles, solving problem styles, epistemological
styles. There was a fruitful discussion concerning the question how the theory of Inge
Schwank concerning individual preferences for cognitive structures fitted into the
general psychological discussion concerning individual styles of cognition.
European Research in Mathematics Education I.II: Group 6 20

3. Cognition and Emotion / Motivation

In general Elena Nardi and Rosetta Zan stressed the importance which emotion plays in
controlling the process of cognition. They pointed out that failure in solving problems
is not only due to a lack of knowledge, but also to the incorrect use of knowledge which
is often inhibited by both, general and specific beliefs: about mathematics, about self,
about mathematics teaching, about social context.

Elena Nardi exemplified research and literature in the field of cognition and emotion
and emphasised the inextricability of emotional and cognitive factors in the formation
of conceptual structures. Current literature suggests links between attitudes towards
mathematics as a lifetime utility tool and the development of conceptual understanding,
between problem solving skills and beliefs about the nature of a mathematical problem,
between development of formal mathematical conceptual structures and beliefs about
the necessity of mathematical proof etc.

Rosetta Zan refined the discussion of affective factors by introducing the distinction
and interplay between beliefs, attitudes and emotions. She revisited the submitted
papers by pointing out how the cognitive analysis presented in them was substantially
incorporating a consideration of beliefs, attitudes and emotions. She subsequently
presented a vista of international works in the area and concluded with potential future
research questions in the area.

4. Cognition and Language

Pearla Nesher gave an overview concerning research in the solving of word problems
under the perspective of language. She pointed out that different wordings may induce
different mental models of the given problem in the pupil’s mind. The study of these
interferences gives fruitful hints for the explanation of pupils misconceptions and
difficulties in mathematics lessons.
European Research in Mathematics Education I.II: Group 6 21

Pierre Luigi Ferrari analysed typical mistakes and difficulties of senior students
while using mathematics language. Pearla Nesher and Pierre Luigi Ferrari put the
question of necessity of special work of students for mastering mathematics language.

5. Cognition and New Technologies

Although there had been a thematic group concerning new technologies the group
leaders had decided that in this thematic group concerning cognition there should be
one discussion concerning the aspect which role the access to new technologies plays
for the development of mathematical cognition and the ways of teaching and learning
mathematics. Tatiana Oleinik pointed out how in the practical teaching process specific
software like Derive and Cabri geomètre foster students’ cognitive growth. She
explained how the use of these software packages can support visual thinking as an
important aspect of mathematical thinking and how it can support an open end
approach in mathematics teaching.

Elmar Cohors-Fresenborg tackled the problem of the session from a more
fundamental point of view. He discussed the interaction between external presentations
and the use of mental models under the perspective that the use of computers enables a
more visual or process-orientated external representation. He pointed out that
according to the distinction between predicative versus functional cognitive structures
the support of the dynamic aspect of using computers may support the more
functional-orientated students while the traditional way of using static formalisations
on paper had supported the predicative ones.

In the following discussion there was a strong debate concerning the question how
far the use of computer in the mentioned way - really from a cognitive point of view -
has to be distinguished from the traditional way of dealing with mathematical
representations on paper. It was pointed out that the specificity of a functional mental
model cannot be represented in a static representation on paper, e.g. neither the concept
of balancing of weight nor the process of switching and its consequence in a circuit.
European Research in Mathematics Education I.II: Group 6 22

6. Mathematical Thinking in Modern Conditions – Situated

In her introduction Elena Nardi pointed out the increasing international interest in
Situated Cognition. Jarmila Novotna then opened up four areas of discussion related to
the use of relevant projects in the mathematics classroom: time considerations, a
definition of a “project”, projects as social problems and the issue of terminology
linked with projects.

Subsequently the group debated on the links between situated cognition and the
development of formal mathematical reasoning. Possible risks were highlighted with
regard to the excessive embedding of the learner’s mathematical reasoning in the
situational structure of a project and alternatives were also brought into consideration.

7. Benefit and Outcome of the Working Group

The discussions in the official group sessions induced in several cases the wish for a
more private and more intensive discussion with the target to understand the
colleagues’ theoretical positions and to go deeper into the different experimental

The discussion showed that there could be different ways of co-operation. One deals
with the question how far behind different wording of theoretical explanations there are
common aspects. A further intensive discussion was to make clear where the
similarities and distinctions between the different theoretical framework lie.

The second possibility of co-operation deals with the idea that an observed
behaviour of mathematical activities can be explained by different scientists using their
different theoretical explanations: Two pairs of scientific eyes should provide a better
picture of the phenomenon which leads to a deeper understanding.

The third aspect for future co-operation concerns the question how to replicate an
experimental design in different countries. This experimental approach is common in
science, but should be more often used in mathematics education.
European Research in Mathematics Education I.II: Group 6 23

Another aspect of further co-operation could be the possibility of adding a specific
situation in an experiment to an additional experimental design from a colleague of a
foreign country. One benefit of such extension could be that both partners have a better
insight into the cultural variances and invariances and the importance of different
curricula or school systems for the analysis and understanding of the mathematical
teaching and learning processes.
European Research in Mathematics Education I.II: Group 5 24

Pier Luigi Ferrari
Dipartimento di Scienze e Tecnologie Avanzate, Università del Piemonte Orientale ad Alessandria corso Borsalino 54,
15100 Alessandria, Italia

Abstract: This paper is concerned with the role of language in advanced mathematical
thinking. It is argued that some behaviors may arise from the application to
mathematical language of some conventions of ordinary language. Grice’s
Cooperative Principle (CP) is introduced in order to explain some episodes that are not
easily accounted for in terms of semantics only. Some examples of (undue) application
of CP to mathematical language are given. It is argued that the application of CP to
mathematical language in problem solving is closely linked to the poor use of
mathematical knowledge and, more generally, to the attitudes and behaviors that
Vinner (1997) names ‘pseudo-analytical’.
Keywords: language, problem-solving, pragmatics.

1. Introduction

The opinion that failure in mathematics is often connected to language is increasingly
popular among mathematics educators and researchers in mathematics education.
From the one hand it is almost common to find some correlation between mathematical
performance and general linguistic competence, from the other hand a number of
studies have pointed out specific obstacles related to mathematical language, in
particular to the interpretation of mathematical expressions and the translation of
ordinary statements into mathematical formalism.

A fair amount of research has been devoted to the analysis of syntax, semantics or
vocabulary of algebraic language in comparison with natural language. For example, a
number of studies have dealt with the so-called reversal error, regarded as an example
European Research in Mathematics Education I.II: Group 5 25

of improper translation. In particular, Bloedy-Vinner (1996) ascribes to the lack of
predicates in mathematical language some improper translations into equations of the
relationships described in the statements of word problems. In another line of research,
Duval (1995) shows that the lack of congruence between representations strongly
affects translations between natural and symbolic language. These streams of research
have no doubt provided convincing interpretations of students’ behaviors and new
teaching ideas. Nevertheless there are behaviors which cannot be accounted for this
way. Let us see two examples. Problems 0.1 and 0.2 have been administered to samples
of freshman computer science students from 1994 to 1997.

Problem 0.1: Is it true that the set A={–1, 0, 1} is a subgroup of (Z,+)?

There are students who claim that A is closed under addition and show that if an
element of A is added to another (different) element of A, the result belongs to A. They
do not take into account x=y=1 nor x=y=–1, the only evaluations which lead to
discover that sum is not a function from A´A to A. In other words, they misinterpret the
definition of subgroup. This happens in spite of students’ knowledge of different
representations (addition tables, …) which point out that an element can be added to
itself. Moreover, they seem aware that each of the variables x, y may assume any value
in A. Most likely, they are hindered by the need of using two different variables to
denote the same number, which does not comply with the conventions of ordinary
language, according to which different expressions (in particular, atomic ones) usually
denote different things.

Problem 0.2: Let m, n be integers such that m×n = 4×6; can you conclude that m=4 and

Affirmative answers are almost frequent. Most of students, if interviewed after the
test, would explain their answer with arguments like “You wrote 4×6 instead of 24;
there must be some reason for this.”

Behaviors like these can hardly be ascribed to the wrong interpretation of specific
expressions or symbols, nor they can be viewed as translation errors; they are often
classified, by university and high school teachers as examples of carelessness or
naïveté, but a closer analysis shows that they can be regarded as the result of improper
European Research in Mathematics Education I.II: Group 5 26

application of conventions and forms of ordinary language (including the less accurate
ones) to mathematics. In both examples students’ behaviors are linked to the
interpretation not only of expressions and statements, but also of the interaction as a
whole and the goals of the interlocutor.

Some ideas and results of pragmatics may help to interpret behaviors like these. I do
not expect that pragmatics can provide a definite theory explaining phenomena that are
specific of mathematical language and theories. What I search for are new perspectives
in order to interpret some behaviors and to start an alternative analysis of mathematical
language. In this paper I try to explain some episodes in terms of Grice’s Cooperative
Principle (1975), according to which a conversation is not a sequence of incoherent
remarks but a cooperative work whose participants recognize a common goal or a
mutually shared perspective. The goal may be fixed from the beginning or may
develop, but there are some behaviors that are improper from a conversational
viewpoint anyway. According to the Cooperative Principle (CP), the contribution of
each interlocutor to the conversation is such as required (at the stage it happens) by the
shared goals and perspectives of the linguistic exchange. This means that, generally
speaking, the amount of information conveyed should be neither more nor less than
required by the situation and that, in most situations, only true sentences are to be
exchanged: for example, it is common to read or hear (in textbooks, school practice, …)
statements like “You cannot write ‘2+2=5’” in place of the more correct “‘2+2=5’ is
false”. Moreover, each interlocutor should be (relatively) pertinent, clear, concise,
neat, …. When interpreting a statement, one generally supposes that the interlocutor is
applying CP. Thus sentences like “Now Mr. X is not beating his wife” (which are often
true) should violate the principle if Mr. X usually does not beat his wife, and the only
interpretation compatible with the presupposition that the interlocutor is applying CP
implicates that Mr. X usually does that. Mathematical language, for different reasons,
often violates CP and students may happen to choose interpretations which are
compatible with the presupposition that the interlocutor is applying the principle even if
they are mathematically inconsistent or meaningless. So the only interpretation of the
use of 2 different variables which is compatible with CP requires that they denote
different objects, and the only interpretation of the use of the expression ‘4×6’ in place
of the more common and simple ‘24’ requires that it should be related to some purpose
of the writer.
European Research in Mathematics Education I.II: Group 5 27

I am not claiming that Grice’s theory is the most suitable contribution from
pragmatics for the analysis of mathematical language (see Caron, 1997, for some
critical remarks). Moreover, most theories of natural language understanding are linked
to different perspectives within cognitive psychology, which cannot be discussed here.
Other frameworks could be used, such as Sperber & Wilson’s theory of pertinence
(1986) or Clark & Marshall’s theory of common ground (1981). But Grice’s theory, for
its simplicity, seems appropriate to carry out an exploratory analysis of some episodes
related to mathematical language from the standpoint of pragmatics. A thorough
analysis of mathematical language from the standpoint of pragmatics is yet to be carried
out; I think it should start from a keen interpretation of use and functions of
mathematical theories (in particular, mathematical logic) and mathematical

In sections 1 and 2 some examples of situations will be presented which could be
accounted for with the help of CP. Students do not apply CP to interpret any statement
or situation. In some cases they seem to interpret expressions according to standard
mathematical semantics. Different parts of the same statement may happen to be
interpreted in different ways, maybe according to their assumed importance or their
exterior aspect. So the parts of the statement that look important may be interpreted
according to students’ mathematical knowledge, whereas the others are interpreted
according to conversational schemes. In section 3 an experiment will be presented
which shows that even slight variations of the statement of a problem as well as of
punctuation or layout may affect students’ resolution procedures as far as students are
prevented from choosing conversational interpretations.

2. The Interpretation of Inequality Symbol

In many contexts students are troubled by statements containing the symbols ‘£’ or ‘³’.
Dealing with questions like “Is it true that 2£100?” answers like “No, it is true that
2<100” are almost frequent. Students’ uneasiness may depend on the manifest
violation of CP: since 2 is clearly much smaller than 100, why to mention the case
2=100, which is obviously absurd? This may be related to the specific English (and
Italian) translation of the symbol ‘£’, which is usually read as ‘less or equal’ (Italian:
European Research in Mathematics Education I.II: Group 5 28

‘minore o uguale’), and then looks like something more complex and expensive (as it
requires the writing or the utterance of 3 words in place of 1) than the plain ‘less’.

The following problem has been given to a group of 34 computer science majors
who had already taken at least 4 mathematics units. The problem has been given within
a Mathematical Logic course, after students had been taught some foundations of
propositional logic (connectives, truth-table semantics, derivations) and in particular
the meaning of words like ‘logical consequence’, ‘incompatible’, ‘independent’ and so

Problem 1.1: Let n be a natural number satisfying both of the following conditions:
(1) If n<100, then n is even (2) If n is even, then n>100

For each of the following statements say whether it is a logical consequence of
conditions (1), (2), or it is incompatible or independent.
(a) n<100 [incompatible] (b) n³100 [consequence]
(c) n is even [independent] (d) n>100 [consequence]
(e) n³80 [consequence]

For each of the statements (a), (b), (c), (d), (e) students could choose among 4
pre-arranged answers: consequence, incompatible, independent, other.

Here we are more interested in the consistency of the answers than in their global
correctness. Let us see the pair (a), (b): answers claiming that one of them is
consequence of the conditions and the other one incompatible or that they are both
independent have been classified consistent, no matter whether correct or not. All the
other answers have been classified inconsistent. Only 17 consistent answers were
found. The other 17 students gave inconsistent answers: most of them claimed that (a)
and (b) are both consequences of or both contradictory with the conditions (1), (2).
Some others claimed that (a) is incompatible (which is correct) and (b) independent
(which cannot hold). If we focus on the pair (b), (e), we find 18 inconsistent answers
(i.e. those claiming (b) incompatible and (d) independent or consequence, or claiming
(b) independent and (d) consequence). If we focus on the pair (b), (e), we find 3
inconsistent answers only.
European Research in Mathematics Education I.II: Group 5 29

Behaviors like these may be regarded as consequences of the interpretation of the
statements according to conversational schemes. Only the links among (b) and (e) are
pointed out even by a superficial interpretation of ‘£’. When interviewed, almost all the
students involved based their interpretations of ‘£’ on common language and
recognized as obvious the transitivity of ‘£’. The same does not work for (b) and (d),
whose comparison requires a subtler interpretation (even from a logical viewpoint) of
symbols ‘<‘ and ‘£’. Even the comparison between (a) and (b) requires to recognize
that each of them is equivalent to the negation of the other (since N is linearly ordered
by £); this involves some knowledge about N which is not explicitly stated in the
conditions (1), (2) and is to be expressed. Maybe the relatively large number of students
claiming that both (a) and (b) are incompatible focused on the conditions (1), (2) only
without using their knowledge about N.

3. A Divisibility Problem

Students can be strongly affected by the wording of the problems they are given. In the
following example, the task required to answer some questions on an unknown number
M, characterized by some conditions, expressed either in Italian language or by
formulas, according to the version.

Problem 2.1.: M is a positive integer satisfying all of the three following conditions:

(Version A)
• there exists an integer p such that M=2p+1;
• there exists an integer q such that M=5q;
• there exists an integer k such that M=7k.

(Version B)
• M is odd;
• M is a multiple of 5;
• 7 is a divisor of M.
European Research in Mathematics Education I.II: Group 5 30

For each of the following statements determine whether it is true, false or other.

(c) M+7 is divisible by 14 [true: M+7 is even and divisible by 7]
(e) M+7 is a multiple of 21 [independent: both 35 and 105 fulfill the conditions]

The pre-arranged answers were: ‘true’, ‘false’, ‘cannot answer, more data required’
and there was a blank space for explanations. The problem has been given to a sample
of 35 freshman computer science students; 17 subjects took version A, 18 version B.

question (c) question (e)
Correct with some suitable explanation 10 7 2 9
Correct with no suitable explanation 7 4 0 5
wrong or missing 0 7 15 4
Totals 17 18 17 18

Tab. I: Answers to problem 2.1

Students with version A performed better on question (c), whereas students with
version B performed better on question (e). As regards question (c), the most difficult
step was seemingly to recognize that M+7 is even. Most of students giving a wrong
answer to question (c) recognized that M+7 is divisible by 7 (mainly by means of some
equation like M=7k) but failed to recognize that it is even. The availability of equations
in version A seemingly helped the students to coordinate divisibility by 2 with
divisibility by 7, even if 7 of them were not able to explain their procedure.

As regards question (e) subjects with version A generally did not try to interpret the
data and design some strategy, but performed calculations or formal manipulations
only; the presentation of data by means of equations might have induced many of them
to believe that the answer should be found by means of formal manipulations and that it
should be a definite one (‘true’ or ‘false’).

Version B induced the subjects to apply everyday-life reasoning patterns; this
version presents some similarity between the external aspect of the conditions and their
mathematical meaning: divisibility by 3 is not mentioned nor implied, as it is
European Research in Mathematics Education I.II: Group 5 31

semantically independent from the conditions. Thus even superficial interpretations of
the statement might lead to correct answers. In both cases the subjects have been
strongly influenced by the version of the problem.

4. Terms and Objects

All the examples in this section regard the relationship between mathematical
expressions and the objects they represent. As already remarked, the use of two
different nouns to denote the same object is a violation of CP. This may trouble students
when it is necessary to give the same value to two or more distinct variables within an
expression. Let us see one more example.

Problem 3.1: Let x, y be divisors of 7; can you conclude that x×y divides 7?

It is not unlikely to find answers like: “Yes, because 7 is a prime, then the divisors of
7 are 1 and 7, then x=1 e y=7 or x=7 e y=1, in both cases x×y=7.” The case x=y=7 is not
taken into account. This problem is even trickier than problem 0.1, because here the
correspondence between names (x, y) and things (1, 7) is 1-1 and, if x=y, the statements
‘x divides 7’ and ‘y divides 7’ become the same, which makes the violation of CP more

4.1 The Role of Problem Presentation in Problem Solving

Similar results are found in problem solving if we break the correspondence between
the clauses which express the conditions to be satisfied and the conditions themselves;
this happens, for example, when more clauses express the same condition. Let us
consider the following problem.

Problem 3.1A: Find out, if possible, a polynomial with real coefficients, of degree 4,
with at least 1 integral root, at least 2 real roots and at least 1 complex, non real root.

An almost frequent answer (by freshman computer-science students) is the
following: “A polynomial like that does not exist, because it should be of fifth degree,
European Research in Mathematics Education I.II: Group 5 32

for it must have 1 integral root, 2 real roots, 2 (conjugate) complex non real roots.”
Students do not realize that the first two conditions (‘at least an integral root’, ‘at least 2
real roots’) designate 3 roots but actually require 2 only. The relevant aspects of this
answer are the following:
• students seem not aware that ZÍR;
• they seem to focus on the (true) fact that the presence of a complex, non real root
implies the existence of another complex non real root (its conjugate).

Explanations based on students’ lack of care or distraction are not satisfactory.
Actually, if we change some features of the problem we find completely different
behaviors. The same students would answer correctly to questions like the following,
asked before or after dealing with problem 3.1A.
• Is it true that ZÍR?
• Are there integers that are not real?
• Is it possible to find an x such that xÎZ but xÏR?

Let us consider a problem of the same kind, given to the same groups of students.

Problem 3.2: Find out, if possible, a polynomial with real coefficients, of degree 2, with
at least 1 integral root and at least 2 real roots.

Usually, this problem is properly solved by a much larger number of students than
3.1A. The number of students correctly solving problems like 3.1A is always less than
50% of the sample (sometimes even less than 20%), whereas problems like 3.2 are
generally solved by more than 80% of the sample.

These data suggest 2 remarks:
• The behavior reported about problem 3.1A seemingly does not depend on the
lack of basic knowledge by students, as they proved to know the same notions if
asked in a different context or form.
• If the notion involved is included within a problem solving context, failure is
mainly caused by the complexity of the problem and the presence of ‘difficult’
steps that draw students’ attention.
European Research in Mathematics Education I.II: Group 5 33

It seems that students focus on the part of the statement they recognize as more
important (related to the intentions of the writer) and interpret it according to standard
mathematical semantics, with no application of CP. For example, the 2nd complex,
non-real root is not mentioned in the statement but the students know, by ‘didactic
contract’ that the theorems on complex roots of real polynomials are part of the
curriculum and are likely to be required. The other parts of the statement seem to be
interpreted according to conversational schemes, i.e. schemes based on common
language which may apply CP less or more consciously. So discouraging students from
applying conversational schemes should produce an improvement of results. This
conjecture has been checked by comparison of problem 3.1A with a modified version,
problem 3.1.B, which is equivalent, but presents a different wording. This procedure
has been arranged for the sake of research and is not suggested as an instructional
method at all. In other words, I am not suggesting that the use of natural language
should be avoided, nor that teachers must choose particular wordings in order to make
easier the interpretation by students.

Problem 3.1B: Find out, if possible, a polynomial with real coefficients, of degree 4,
satisfying all of the following conditions:
• at least 1 of its roots is an integer;
• at least 2 of its roots are real;
• at least 1 of its roots is a complex non real number.

The two presentations are similar even as regards vocabulary. Version B has been
written in order to discourage interpretations based on conversational schemes, by
means of punctuation, layout and the repetition of part of the sentences (“at least … of
its roots is/are …”). This could have marked the difference of the situation from a
everyday-life one, so that students were prevented from applying CP.

4.2 Some Results

The two versions were given to 25 freshman computer science students, at the end of 1st
semester, after taking some 30 hours of introductory algebra. 13 students took version
A, 12 took version B. The two samples proved equivalent, as they performed almost
European Research in Mathematics Education I.II: Group 5 34

identically in all the other problems of the test.

Correct answer Wrong answer No answer
problem 3.1A 4 (31%) 7 (54%) 2 (15%)
problem 3.1B 7 (58%) 2 (17%) 3 (25%)

Tab. II: Answers to problem 3.1

These data are not conclusive. Further research is needed to corroborate the
interpretation suggested. Nonetheless, the analysis of protocols shows that students
with version B paid more attention to the interpretation of the data, by means of Venn
diagrams or explicit remarks like ‘ZÍR’ and so on. Students with version A, including
those providing a correct answer, generally did not pay much attention to the
interpretation and the representation of the data.

After the test all the subjects were interviewed by senior mathematics majors. All
the subjects giving a wrong answer ascribed their errors to lack of care. All the students
have been shown both versions and have been asked if they believed that the
presentation could have affected their answers. 21 students, out of 25, claimed that the
wording of the statement of the problem did not affect their answers.

The outcome of the test and the interviews suggest two remarks:
• The wording of the problem did affect the answers to some extent.
• Students were not aware of this effect.

5. Discussion

The findings reported in sections 1, 2 and 3 show that the conversational interpretation
of mathematical statements may account for a number of behaviors. But phenomena
like these do not depend only upon language. They involve students’ attitudes toward
mathematics and mathematical tasks as well. Then I am not claiming that the
misinterpretation of the texts of the problem is the only reason for the errors reported.
The point is that often students do not use their mathematical knowledge in order to
solve a problem. Some answers to problem 1.1 are examples of a behavior like that, as
European Research in Mathematics Education I.II: Group 5 35

students do not use the obvious fact that, if n is an integer, ‘n<100’ is true if and only if
‘n³100’ is false. Another example is problem 3.1A, where the obvious fact ‘ZÍR’,
which could easily avoid the misinterpretations reported in section 3, is overlooked. In
other words, the use of knowledge should increase effectiveness but also stability of
resolution procedures. More examples at this regard can be found in Ferrari (1996,
1997). Roughly speaking, it seems that students apply CP in some situations and their
knowledge in others. The criteria they follow to choose whether to apply conversational
schemes or more specific knowledge are related to what Vinner (1997) names
pseudo-analytical behavior. Often students do not read the statement of the problem
with the goal of recognizing the mathematical context they are asked to work within
and to re-construct the knowledge that is required; they simply read the statement (or
only a portion of it) in order to find verbal clues that could tell them which behavior
(maybe out of pre-arranged set) they are expected to activate. This could explain why
students’ behavior depends upon the wording of the statement so strongly: some words
and layouts suggest them to apply their mathematical knowledge (or
pseudo-knowledge), activating a sort of script selected from their experience in
mathematics. From this viewpoint, interpreting the purposes of the interlocutor
becomes far more important than recognizing the mathematical ideas involved in the
problem. This obviously amplifies the effects of the undue application of CP (or other
improper interpretation schemes) to mathematical statements. I think that the link
among linguistic obstacles and pseudo-analytical behavior is a fundamental one:
should students actually read the statements with the purpose of understanding their
meaning, a weaker dependence on the wording would be found. In most cases students
possess some basic linguistic skill, but do not use them when reading the statement of a
problem, as they search for verbal clues only, and so even a slight change may affect
their interpretation.

I am not suggesting that ordinary language is a source of errors or may induce
non-formal mathematical behavior. On the contrary, I think that the use of ordinary
language when doing mathematics is unavoidable and a powerful tool to understand
and communicate. But inaccuracy in the use of ordinary language, and the lack of
knowledge of functions and usage of mathematical language (including symbolism but
also ordinary statements when interpreted according to mathematical semantics) often
go together with pseudo-analytical behavior. From the one hand, people working in the
European Research in Mathematics Education I.II: Group 5 36

pseudo-analytical mode do not try to interpret statements according to their
mathematical knowledge and thus it is natural for them to apply conversational
schemes; from the other hand, people who cannot interpret mathematical statements
reason are prevented from using their knowledge (if any) and compelled to search for
verbal clues and to adopt pseudo-analytical behaviors.

These effects are increased by the practice of working within one representation
system only, which make difficult to recognize equivalence or analogies between
problem situations or to avoid obstacles. Sometimes students know different
representations but cannot use or coordinate them (see Duval, 1995). Some answers to
problem 2.1 are examples of lack of coordination of representation systems.

A thorough discussion on the instructional methods that could overcome these
obstacles is out of the reach of this paper. Experience shows that students with good
linguistic competence and accustomed to use language in scientific contexts too (not
only to use language to show they master language) generally can easily overcome
these obstacles. Working with multiple representations, with the help of technology, is
another factor that can improve understanding. But any effort will be hopeless if we do
not remove the factors which induces students to try to solve problems with no use of
knowledge. Most likely it is neither possible nor desirable, to completely inhibit
pseudo-analytical behaviors (for a discussion on this aspect see Vinner, 1997). But
maybe it is possible to prevent them from hindering any further understanding. This
requires deep changes in instruction and in assessment methods.

6. References
Bloedy-Vinner, H. (1996). The analgebraic mode of thinking and other errors in word problem
solving. In A.Gutierrez & L.Puig (Eds.). Proceedings of the 20th Conference of the
International Group for the Psychology of Mathematics Education, vol.2, 105-112.
Brousseau, G. (1986). Fondements et méthodes de la didactique des mathématiques. Recherches en
didactique des mathématiques, 7(2), 33-115.
Caron, J. (1997). Psychologie cognitive et interactions conversationnelles. In J.Bernicot,
J.Caron-Pargue & A.Trognon (Eds.), Conversation, interaction et fonctionnement cognitif,
221-237. Nancy: Presses universitaires.
European Research in Mathematics Education I.II: Group 5 37

Clark, H.H. & Marshall, C.R. (1981). Definite reference and mutual knowledge. In A.K.Joshi,
B.Webber & A.Sag (Eds.), Elements of discourse understanding, 10-63. Cambridge:
Cambridge University Press.
Duval, R. (1995). Sémiosis et pensée humaine. Bern: Peter Lang.
Ferrari, P.L. (1996). On some factors affecting advanced algebraic problem solving. In A.Gutierrez &
L.Puig (Eds.), Proceedings of the 20th Conference of the International Group for the
Psychology of Mathematics Education, vol.2, 345-352.
Ferrari, P.L. (1997). Action-based strategies in advanced algebraic problem solving. In E.Pehkonen
(Ed.), Proceedings of the 21th Conference of the International Group for the Psychology of
Mathematics Education, vol.2, 257-264.
Grice, H.P.(1975). Logic and Conversation. In P.Cole & J.L.Morgan (Eds.), Syntax and Semantics -
Speech Acts, 41-58. New York: Academic Press.
Sperber, D. & Wilson, D. (198). Relevance: Communication and cognition. Oxford: Blackwell.
Vinner, S. (1997). The Pseudo-Conceptual and the Pseudo-Analytical Thought Processes in
Mathematics Learning. Educational Studies in Mathematics, 34, 97-125.
European Research in Mathematics Education I.II: Group 5 38

Emanuila Gelfman, Marina Kholodnaya
Department of Algebra & Geometry, Tomsk Pedagogical University,
Komsomolskii Prospect, 75, 634041Tomsk, Russia
Laboratory of Psychology of Abilities, Institute of Psychology of Russian Academy of Sciences,
St. Yaroslavskaya, 13, 129366 Moscow, Russia

Abstract: Ways of information coding are subjective means with the help of which the
surrounding world is reproduced in individual experience. Mastering mathematical
concepts presupposes solution of two didactic tasks: firstly, including in the process of
teaching three ways of information coding: verbal, visual and sensual - sensory with
consideration of certain requirements to introduction of each of them; secondly,
organisation of self-transference in the system of these three ways of information
coding. Such activity in the process of work with concepts should lead to success of
individual intellectual behaviour.
Keywords: ways of information coding, cognitive competence, individualization of
intellectual activity.

1. Introduction

Ways of information coding are subjective means with the help of which the
surrounding world is reproduced in individual experience and which ensure
organisation of this experience for future intellectual behaviour.

Principal ways of information coding were described in the works of George Bruner
(1971, 1977). Bruner speaks of existence of three principal ways of subjective
presentation of information: in the form of object actions, visual images and linguistic
signs. Development of a child’s thinking takes place in the course of a child’s mastering
these three ways of subjective information coding which are in relations of mutual
influence and interaction. Analogous idea that the work of a thought is ensured by three
European Research in Mathematics Education I.II: Group 5 39

“languages” of information processing - linguistic, image - bearing and
tactile - kinaesthetic — was repeatedly expressed by L. M. Vekker (1976).

So, in our opinion, the informational exchange between a person and environment is
ensured by three principal forms of experience: 1) in the form of verbal signs (verbal
way of information coding); 2) in the form of images (visual way of information
coding); 3) in the form of sensual impressions (sensory way of information coding),
(Kholodnaya, 1997).

It is necessary to underline that the problem of the ways of information coding is not
completely identical to the problem of forms and representations of information.
Information may be represented to a person in various forms: by means of somebody’s
speech, by means of illustrations in a text-book and by a text itself, by means of
computer ambience, by demonstration of real objects and so on. But any information,
irrespective of its source, is transferred to individual subjective experience by three
interacting psychic “channels”. This basic cognitive mechanism, namely: mechanism
of reversible mutual transference in the system of three ways of information coding
influences two basic lines of a child’s intellectual development. Its formation
determines, firstly, growth of conceptual competence due to integration of different
forms of experience and, secondly, growth of individualisation of intellectual activity
due to revealing individual intellectual styles.

Within the frames of this paper the following questions of algebra teaching will be
1) in what form should verbal, visual, sensual ways of mathematical information
coding be introduced, as well as how to organise the transference of experience from
one form to another one;
2) in what way, on the basis of mutual transference of information in different forms of
experience, can we ensure revealing and further forming of students’ individual
intellectual styles?

In our opinion, these questions may be solved by means of organisation of school
texts. Such work was performed in the project “Mathematics. Psychology.
Intelligence” in the frames of “enriching model” of teaching.
European Research in Mathematics Education I.II: Group 5 40

2. Ways of Information Coding as One of the Conditions of
Growth of Conceptual Competence

Researches of conceptual thinking showed that the process of concept functioning
presupposes simultaneous participation of three forms of experience — verbal,
including the usage of words of the native language and artificial signs, visual and
sensory. Accordingly the conclusion was made that conceptual thought is the result of
reversible mutual transferences in the system of three “languages” of information
processing (Vekker, 1976; Kholodnaya, 1983).

In this case, it is important, in our opinion, to underline the following. On the one
hand, concept is presented as an unit of knowledge which exists objectively and which
a child learns in the process of education. On the other hand, a concept is formed (made,
summed up) within individual mental experience, acting as conceptual psychic

Taking into account what has been said above we may schematically represent the
process of forming conceptual structure in the following way:

Sign definition

as an element
of mental experience

Object Sensual
(real or ideal) impressions

Fig. 1: Correlation of verbal, visual and sensory forms of experience
in the process of concepts forming

So, when we comprehend anything at the conceptual level, we give it verbal
definition, we have its image in our mind and we spontaneously feel it.

The problem of taking into consideration the psychological nature of the process of
formation of concepts is rather important in educational process. Very often, traditional
European Research in Mathematics Education I.II: Group 5 41

teaching, making the words (signs, formulas, symbolic expressions) nearly the only
means of intellectual communication with a child, ignores the key meaning of the other
two, equally important for development of intellectual abilities of children to
accumulate knowledge about the world — through action and image. However,
without usage and proper organisation of sensory (including active) as well as visual
experience of a child, complete mastering of the meaning of signs and symbols (at the
level of comprehension of concepts) becomes difficult. Language “codes” run idle,
touching only superficial layers of a child’s ideas of the world.

So, forming concepts “inside” a child’s mental experience, that is growth of his/her
conceptual competence, presupposes taking into account one of the basic cognitive
mechanisms of intellectual development — mechanism of intercoordinated
functioning of these three “languages” of information coding. Accordingly, mastering
mathematical concepts requires solution of two didactic problems:
1) including in the process of education three ways of information coding — verbal,
visual, sensual - sensory — with consideration of certain requirements to the
introduction of each of them;
2) organisation of reversible mutual transference of information in the system of these
three ways of information coding.

Verbal way of information coding may be introduced by active participation of
students in formulating definitions, creating a formula, comparing different definitions
and records of mathematical expressions. Students should also with the help of a word
describe characteristics of objects and links between them.

Let us give an example of such work with the theme "Monomials". Students are
given the following task:
Task 1. (Gelfman 1998, pp. 76-77).
Answer the following questions in writing:
a) How many months are in t years?
b) How many hours do n minutes make?
c) How many cubic centimetres are in m cubic metres?
d) How many minutes are in m days, in n hours?
e) What is the area of a figure made of three equal rectangles with sides a and b ?
f) What is the volume of a body made of five equal parallelepipeds with edges a, b and a ?
g) What is the area of a square with side c ?
European Research in Mathematics Education I.II: Group 5 42

h) How many metres are in a kilometre?”

Task 2. (Gelfman 1998, p. 77).
Record the following algebraic expressions:
a) length of a circumference, an area of a circle of radius r ;
b) product of a variable of the 5th power x and a variable of the 4th power y , divided by six;
c) product of tripled product of a square of variable x and variable y by doubled product of
cubes of the same variables.
d) product of variable x in the 5th power by variable y in the 4th power;
e) a half of the above-given expression;
f) the quadruplicated above-given expression.

After doing this task, students may check the results of the work of changing the
natural language for the language of mathematics by means of the expressions which
are written on the blackboard in a classroom:
50 ; a 2b ; 4x 3b 2y 4 ; (-2)a 3x 2yc 4 ; 12t ;
x 5y 4 ; c2; 3ab ; ; 5a 2b ;
6x5y4 ; 60n ; 24·60m ; 3x2y·2x3y3 ; 103 ;
1 5 4 æ1 ö
25a 3bc·(0,2)a 2cb 2 ; x y ; 4 · ç x 5y 4 ÷ ;
2 è2 ø
1 2
4 a x·(-4)axyc 4 ; pr 2 ; 2pr ; 106·m .

After this work, the students notice that all these expressions should get a new name
as the students have not been confronted with such expressions before. The teacher says
that all these expressions have one name in common "monomial". Then the students are
given the following task:
Analyse the structure of all these expressions and try to define a monomial. What is your
opinion, which of the following statements may be the definition of a monomial? Which of
them can not be the definition of a monomial?
1. Product of some powers, the foundations of which are variables or numbers.
2. Algebraic expression, containing product of variables.
3. Algebraic expression, containing only operations of multiplication and raising to a power.
4. Algebraic expression, containing a variable.
5. Algebraic expression, which is a product, the multipliers of which may be numbers, one or
a few variables, each of which is raised to a certain power.
European Research in Mathematics Education I.II: Group 5 43

6. Algebraic expression, which is a product, the multipliers of which may be powers of
variables and numbers.”

Doing this task, students should point out the characteristics of the notion
"monomial" and, if they disagree with any definition, they should give their arguments.

Then we give some more tasks for students to work in more differentiative and
variative ways with records of monomial and its variants. Two examples of such work
are given below.

Task 3. (Gelfman 1998, p. 86).
Fill in the table

Power of Standard Example of a
Monomial Coefficient Letters
a monomial form similar monomial

5a 5b 5 5 … 5 5a 5b 5 – 2a 5b 5

3 æ 5 3ö
ax ç - ax ÷ … a 2x 4 … … …
4 è 6 ø

… 6 x 5y 4z 3 12 … …

…a 3b(2,5)a 2b 2 6 … 8 … …

Task 4. (Gelfman 1998, p. 85).
Find the numerical values of monomials:
3 æ 5 ö1
1) ab 2x ; 2) 0,5abxb ; 3) 10b 2ax ; 4) – 10axb 2 ; 5) - axç - ax 3 ÷ b 2ax ;
4 è 6 ø2
if a = 2 ; b = –3 ; x = 4 .
The task may be easily done, if you see, that …

It is very useful to give students ability to show their creative abilities in using verbal
ways of information coding. So, working with the same theme "Monomials", we may
give students the following task:

Task 5.
You are a TV-showman, and you should introduce Mr. Monomial to interviewers. Submit
your questions and illustrations for the programme to the editor (a teacher).
European Research in Mathematics Education I.II: Group 5 44

Visual way of information coding presupposes, as one of the variants, usage of
normative images. A table of orders, numerical ray, numerical axis, intercept, graph of
a function, areas of figures and the like are referred to such images.

Special work should be conducted to make these images generalised and dynamic.
So, for example, many teachers use the table of orders only for introduction of natural
numbers but this image is not developed the future. But it is very useful to develop this
image while studying operations with natural numbers and with decimal fractions.
Then, the image will help students to be active in formulating rules of operations with
decimal fractions, because he knows how to act in the table of orders to the left of
decimal comma, so he would project the actions to the right of decimal comma. Here
are examples of development of work with table of orders:

Task 6. (Gelfman 1997, p. 77).
Read and record by words and digits:

Thousands Hundreds Tens Units Tenths
6 , 6
One plus nine tenths
0 , 4
One hundred plus seven tenths
1 2 3 6 , 0

Special role is given to individualised images of students themselves. So, speaking
of the ways of solving rational equations, a student suggests using schematic
representation of these images in the form of a bunch of keys (Fig. 2). Later we included
the following picture of the bunch of keys in the text-book (Gelfman 1996b, p. 207).

One more example of visual way of information coding. In the process of looking
for definitions of multiplication of different numbers, monomials, polynomials, it is
very useful to give the task of defining areas of rectangles and squares. It is important
for students to see development of these images in the process of learning different, but
closely connected, products. Here is one of the tasks, which serves this purpose.
European Research in Mathematics Education I.II: Group 5 45

Fig. 2: Student’s suggestion for schematic representation
of the ways of solving rational equations

Task 7. (Gelfman 1996b, p. 91).
Look at the drawings and restore all the algebraic expressions which have been omitted in
the recordings.

I 10x2 15x 5x
2x 3
a b
5x(2x + 3) = 10x2 + 15x
… = a2 + 2ab + b2

2x 3
8x 12 4
II b a
10x 2
15x 5x
(5x + 4)(2x + 3) = … a2 - b2 = …

The same task is one of the examples of work for transition of one form of
information coding into another one. Such work is necessary for forming any
mathematical notion.

Sensual-sensory way of information coding may be introduced through
objective-practical motivation, performing objective actions with some material
objects, using sensual evaluations (bulky-compact, strict-chaotic, beautiful-ugly and so
European Research in Mathematics Education I.II: Group 5 46

on). So, for example, to motivate usefulness of bringing monomial to standard form we
say to students:
Pay attention to the fact that a standard form of a monomial is much simpler than the
original one. How many, for example, different types of cards would you prepare for
monomials, which we have got in the task?

3. Ways of Information Coding as One of Conditions of
Individualisation of Intellectual Activity

In the process of learning mathematical concepts, different students have different
correlation of three basic ways of information coding. One or another form of
experience may be more predominant; to one student it is necessary to explain some
material in words; to another student it is necessary to show and to a third one — to give
the possibility to experience (including one’s own actions). On this basis individual
cognitive styles, by which we understand individualised ways of studying reality, are

Formation of mechanism of reversible mutual transformation of three ways of
information coding gives the possibility of organising an educational text in such a way
as to represent ways of teaching students with different styles of cognition (different
casts of mind).

In particular, predominance of verbal form of experience may find its expression in
verbal-analytical, verbal-categorical, verbal-algorithmical cognitive styles, visual form
of experience — in figurative-illustrative and figurative-modelling cognitive styles,
sensory form of experience — in active-practical, associative-game and
sensual-intuitive cognitive styles.

The work with students in our text-books is organised so that students with different
cognitive styles could try different strategies of education which they could use.

So, for example, students were given the task to write the contents of a paragraph
( a + b ) 3 = a 3 + 3a 2b + 3ab 2 + b 3
European Research in Mathematics Education I.II: Group 5 47

Students with verbal-analytical cognitive styles gave, for example, such tasks for
this paragraph:
Raise to the third power binomials a + b and a - b . Read the identities, you’ve got.
Describe the right parts of the formulae: point out the number of monomials, their signs,
coefficients, exponents.

Students with verbal-categorical cognitive styles suggested carrying out an
"You already know the identities
2 2 2. 3 3 2 2 3
( a + b ) = a + 2ab + b ; ( a + b ) = a + 3a b + 3ab + b .
Can you open the brackets in the expression ( a + b ) ? What do you think, how many
members will be in the expression you’ve got, how will the exponents work, what will be
values of the coefficients?" "Get formula of transformation of cube of trinomial ( a + b + c )
into multinomial."

Tasks of students with verbal-algorithmic cognitive style were like that:
“Compose algorithm of getting multinomial from cube of sum”, “Consider two solutions and
mark basic steps in them”.

Tasks of the type:
“Try to compose domino, lotto, labyrinth, in which identities
( a ± b ) 3 = a 3 ± 3a 2b ± 3ab 2 ± b 3 may be used”,
“Spot the mistakes …”
were suggested by students with associative-game cognitive styles.

Students with figurative-illustrative style gave such tasks:
“Look at the drawing … . Explain it with the help of formula …”;
“Make up a scheme of formula of cube of binomial”.
Students with verbal-demonstrative cognitive styles worked with such tasks:
“Prove the identity ( a + b ) 3 = a 3 + 3a 2b + 3ab 2 + b 3 ” ;
“Fill in the blanks in the identities …”; “Complete the expressions so that they may be
transformed with the help of formulae …”.

Students with sensual-intuitive cognitive style gave the task:
Consider tasks … Try to make an advertising slogan in the name of the task itself. For
example, “Meeting the situations in which it’s expedient to use identities”.
European Research in Mathematics Education I.II: Group 5 48

4. Conclusion

Intellectual up-bringing of a child in the process of school education presupposes, on
the one hand, raising productivity of his/her intellectual abilities (in particular, growth
of cognitive competence, which is directly related to success of individual intellectual
behaviour. On the other hand, strengthening individual originality of his/her cast of
mind (in particular, formation of individual cognitive style, which provides
effectiveness of individual intellectual adaptation to the requirements of environment).
Solution of these two tasks is connected with working out comparatively new
technologies of teaching, oriented at actualisation and enrichment of basic components
of mental experience of every child (including the level of his/her cognitive,
metacognitive and intentional experience). Within the frames of this paper we tried to
discuss possible variants of consideration of one of components of cognitive
experience, namely: ways of information coding. It allows, in our opinion, to raise the
quality of students’ mathematical knowledge and to individualise the process of
teaching mathematics.

5. References.
Bruner, G. (1971). On cognitive development. In Bruner G., Olver R., Greenfield P. (Eds.) Research
on development of cognition, 25-98. Moscow: Pedagogics.
Bruner, G. (1977). Psychology of cognition. Moscow: Pedagogics.
Vekker, LM. (1976). Psychic processes. Thinking and Intelligence. Leningrad: Leningrad Univ.
Publishing House.
Kholodnaya, M.A. (1997). Psychology of Intelligence: Paradoxes of Research. Tomsk: Tomsk Univ.
Publishing House.
Gelfman, E. et al. (1996a). Identities of Abridged Multiplication. Tomsk: Tomsk Univ. Publishing
Gelfman, E. et al. (1996b). Algebraic Fractions. Tomsk: Tomsk Univ. Publishing House.
Gelfman, E. et al. (1997a). Decimal Fractions in the Moominhouse. Tomsk: Tomsk Univ. Publishing
Gelfman, E. et al. (1998). Making Friends with Algebra. Tomsk: Tomsk Univ. Publishing House.
European Research in Mathematics Education I.II: Group 5 49

Emanuila Gelfman, Raissa Lozinskaia, Galina Remezova
Department of Algebra & Geometry, Tomsk Pedagogical University, Komsomolskii Prospect, 75,
634041Tomsk, Russia

Abstract: Ability of using different ways of work with a text-book is one of the most
important intellectual characteristics of student’s personality. It may be formed by a
teacher and by a student himself by means of specially designed tasks.
Keywords: -

To make a student successful in his mathematics studies a student should not only have
some particular skills but some general skills as well. One of the most important skills is
the skill of information processing, in particular, to work with educational literature. An
outstanding Russian mathematician V.M. Glushkov writes: “There is the necessity of
information processing in every sphere of human activity. Namely, a translator, an
economist, a mathematician and, even, a poet has to process information. Information
processing is what we call human intellectual activity”.

The problems of work with school-texts are treated in the works of Bell (1987),
Granik (1988), Getsov (1989) and others.

The group of the authors of the project “Mathematics. Psychology. Intelligence”
(headed by Prof. E. Gelfman), wrote a series of text-books for students, aged 10-15.
The main idea is that a text-book should perform a role of an intellectual self-instructor,
where a student works directly with a text-book and that gives him the possibility of
finding the most suitable and convenient for him way of mastering school material.
Such an approach presupposes that a student should be, first of all, taught how to use a
text-book and this instruction (in this or that form) is given in every text-book of our
European Research in Mathematics Education I.II: Group 5 50

First of all, the books themselves teach students different ways of information
processing. They may show to students, how, for example, one and the same material
may be represented as a story, a dialogue, a glossary, a system of tasks.

Our text-books also teach students, aged 13-15, how to use reference books. Every
text-book is supplied with special sections, entitled “Glossary”, “Reference book”,
“Samples of solving problems”. There are also special tasks which stimulate the work
with reference-books and glossaries, samples of problem-solving not only from a given
textbook. Students are given tasks to make some entries into glossaries and
reference-books, to make their own samples of problem-solving. Here are two
examples of such type of work:

How do you usually act when you are confronted with a new problem? Do you try to find
the solution independently? Do you ask help from a teacher or a friend? Do you read
special books? Do you turn to reference books? Let’s try to turn to reference book on
mathematics. Let’s take, for example, “Reference book on mathematics”, written by
A.G. Tsipkin. We shall look for the problem requires in the table of contents and we shall
look into the index of problems. Though, we don’t know yet, what word should be looked
for in this index. First of all let’s turn to the table of contents. These is point 5.3, which is
entitled “Quadratic equation”. Here the equation ax2 + bx + c = 0 is given, which is called
quadratic. Equation 2x2 – 53x + 300 = 0 looks very much like it. We have only to realise
that a = 2, b = –53, c = 300. So, what is said in the reference book about solution of
such an equation? … . (Gelfman, 1997).

Reduce fractions to a common denominator:
x y a b c
a) and b) 2
; ; ;
x+y 2x a -1 a + 1 a -1
m n p
c) 2
; 2
; 2
a + 2a + 1 a - 2a + 1 a -1
c2 a5 4 1
d) and ; e) and .
12ab 2 8b 3c 2 a -b b-a

Try to formulate the rule for reducing algebraic fractions to a common denominator. If you
have problems, turn to a reference-book. (Gelfman, 1996a).

We give students tasks to write summaries of a text, make advertisements and
anti-advertisements of some methods of solution of problems of a certain type or of
European Research in Mathematics Education I.II: Group 5 51

some notions. We also teach them how to write synopses, notes, how to present
historical data or how to construct a text. Here are some examples.

Adding of Integers

with identical signs with different signs and with different signs and with a zero:
different moduluses: identical moduluses:

3 + 2 = 5; –5 + 2 = –3; –5 + 5 = 0 –5 + 0 = –5;
–5 + (–2) = –7 +5 + (–2) = 3 0+1=1

Common sign Sign of a “strong” Just simply
and the sum of summand and differ- we get a zero
moduluses ence of moduluses

Using Malbina’s abstract answer the question:
1. In what cases you have to:
a) to add natural numbers
b) to subtract natural numbers for getting the sum of integers?
2. What number should be added to –5 to get 0; –5; +5 ? (Gelfman, 1996b).

Write a short composition on one of the problems given below:
“Application of Viet’s theorem”, “Assumptions which may be driven from Viet’s theorem”,
“Roots of quadratic equation and Viet’s theorem”, “What new material have I learnt thanks to
Viet’s theorem?”, “Around Viet’s theorem”. (Gelfman, 1997).

The following forms of work with texts are given in our textbooks:
1. Page by page text analysis
2. Making plans of texts
3. Making up questions to a text
4. Reviewing
5. Foretelling the contents of a paragraph by its title
6. Conducting role-games
7. Assessment of the tasks given in a text-book
8. Rendering a text from a student’s or a character’s name

and so on.
European Research in Mathematics Education I.II: Group 5 52

Sometimes we devote the entire period to work with books: having such classes as:
conferences, consultations, competitions for the best lecture, text or story.

As our experience shows such work at mathematics lessons helps students to be
more successful not only in mathematics but in other school subjects as well.

1. References
Bell, A., et al. (1987). Awareness of Learning, Reflection and Transfer in School Mathematics.
Teachers’ Handbook. ESRC Project: R000-23-2329. Nottingham: Shell Centre, University of
Nottingham in collaboration with teachers and schools of Wakefield LEA/Nottingham, Derby
and Leicester LEAs and supported by the Economic and Social Research Council.
Granik, G. (1988). When a book teaches. Moscow: Pedagogics.
Getsov, G. (1989). Work with a textbook: rational devices. Minsk: Polimya.
Gelfman, E. et al. (1996a). Algebraic Fractions. Tomsk: Tomsk Univ. Publishing House.
Gelfman, E. et al. (1996b). Positive and Negative Numbers in Pinnocchio’s Theatre. Tomsk: Tomsk
Univ. Publishing House.
Gelfman, E. et al. (1997). Quadratic Equations. Tomsk: Tomsk Univ. Publishing House.
European Research in Mathematics Education I.II: Group 5 53

Marie Kubinova1, Jarmila Novotna1, Graham H. Littler2
Faculty of Education, Charles University, M.D.Rettigove 4, 116 39 Praha 1, CR.
University of Derby, Mickleover, Derby DE3 5GX, UK

Abstract: The teaching of mathematics in the Czech Republic has traditionally been of
an ‘instructive’ nature. This teaching strategy has alienated most students in
mathematics because they have been expected to use the skills they have learnt without
understanding the underlying concepts. They have not been able to appreciate the
usefulness of mathematics nor get enjoyment from the subject. From our own teaching
experience, by using projects and mathematical puzzles we have found that our
students have gained the necessary understanding, enjoyed their work and developed
other important attributes such as the ability to conjecture, to work systematically and
to communicate. In this paper we set out the benefits of using this practical approach to
the teaching of mathematics together with an analysis of the processes necessary to set
up such teaching strategies in schools etc.
Keywords: investigative methods, understanding.

1. Introduction

The prime source of our interest for the use of investigative teaching methods in
mathematics education was the long-term experience of one of the authors with her
own direct teaching of mathematics, whom did not fit the traditional concept of
teaching in Czech schools and the experience of our English colleague who has been
using investigative approaches in school and university since the early 1960s. This
conception is based on
European Research in Mathematics Education I.II: Group 5 54

• the teaching of school subjects separately (even of individual fields such as
algebra or geometry in mathematics),
• instructive delivery of complete knowledge,
• repetitive practice of skills.

The main consequences of it are:

• The emphasis in teaching is put on the content, not on the student’s learning.
• Knowledge and skills gained by most students are not permanent and students
are not ready to use them and to find new or missing information.
• Students are unable to use their mathematical skills in informal examples and
• Students are often afraid of mathematics because of the abstract way in which it
is taught; they are stressed and consequently their performances do not
correspond with their abilities.
• Students (particularly the less-able ones) never have the possibility to feel the
pleasure of “discovering” something new and of achieving their own potential.
• Students are not able to plan, develop strategies, work independently or react
flexibly in concrete situations.
• Students are lacking in creativity, flexibility, critical thinking, the ability to
reason meaningfully, the ability to think procedurally, to understand
• Students lack initiative, self-reflection, self-confidence, and ability to overcome
• Students’ skills work co-operatively in groups and to communicate the results of
their work are not developed.
• The empathy is not developed.
• There is a lack of enjoyment in mathematics.
European Research in Mathematics Education I.II: Group 5 55

2. Our Approach to Projects

We chose an investigative teaching strategy because it is one of the possible ways of
eliminating the problems that the traditional authoritative and instructive teaching
produces for a student. It was mainly this method’s naturalness with its emphasis
focussed on the student and the complexity within projects that made us use it. In full
agreement with (Grecmanova & Urbanovska, 1997), (Littler & Koman, 1998), when
preparing/using investigative teaching we take advantage of the following facts:

1. The student inclines to relate all his/her activities with his/her specific need or
function in every moment.
2. The student’s understanding of reality is gathered in natural way from his/her
autonomous experiences, by direct contact with the world of mathematics, in the
form of experimenting and investigating . This is what J.A. Comenius has already
stated (Comenius, 1631): “It is necessary to proceed from concrete, because it is the
only item that is familiar to the student.”
3. The student does not learn already complete and/or abstract knowledge by means of
algorithms, but constructs this knowledge and during any solving procedure he/she
shapes their knowledge to enable the various parts to connect. Therefore, as
a necessity during the projects construction, we put great emphasis on the following
demand: each project should offer the student a real problem in a relevant context,
a challenge for them to create or construct their personal strategies for solutions and
understanding of mathematics. For us it means that our student’s learning processes
become cognitive instead of only being pedagogic.
4. The student is able to develop his/her cognitive abilities by practical experience and
cognition being the basis of understanding. In a non-instructive way we teach the
student something that we consider to be more important than skill acquisition itself,
namely methods of cognition.
5. The student’s free choice of direction of cognition and better opportunities to
participate in the project processes are strong motivating factors. The ‘real life’
possibilities of making a true or false conjectures add to the student’s motivation and
develop perservence.
European Research in Mathematics Education I.II: Group 5 56

6. Asking the student to work beyond their ability does not occur and his/her individual
needs are prerequisite.

3. What Is a Project

A general definition of the concept of project/investigative teaching is not given here.
There is no agreement on a definition in literature. Different approaches to the
definition and classification have been occurring in literature from the very beginning
of investigative teaching (Dewey, Kilpatrick or Vrána, PÍíhoda in Czech books) until
today. At present investigative teaching strategies are having a renaissance (IHME,
1996), (Littler & Koman, 1998), (Koman & Ticha, 1997).

One of the definitions which most closely agrees with our thinking is that written
specifically for teaching of mathematics and given in (Kubinova, 1997):

A project in mathematical education means any independent work produced
by a student or a group of students that leads to the active solving of a problem
connected with a mathematical concept or mastering a skill. Its main feature is
that a student can decide independently how and in which order he/she will
solve the tasks necessary to successfully cope with the project.

In our mathematics teaching we distinguish three basic types of projects:
inter-disciplinary, inner-disciplinary and mathematical puzzles (tab. I).

1a) 1b) 2)
Mathematical puzzles
Inter-disciplinary projects Inter-disciplinary projects
Developing logical thinking,
from different disciplines. from one discipline.
strategies for winning or ‘not losing’,
giving repetitive work of weak process.
The student experiments, discovers, conjectures,
Clarity of aims and final goal.
proves, verifies, … relations with the project.

Tab. I
European Research in Mathematics Education I.II: Group 5 57

Note: In 1, the projects can often be attempted by students of differing mathematical
ability. Their final work which is presented should indicate the level of their
mathematical thinking (Kubinova, 1997), (Littler & Koman 1998).



• Dandelion seeds are scattered by the wind. Devise a long-term experiment (over
the spring semester) which would provide you with data to show in which
direction you are likely to find next years flowers.
• In one of his stories, Jack London describes his trip by sledge pulled by seven
dogs from Skagway to the camp. The first day, the sledge run at full speed, but
the next day, some dogs ran away with a pack of wolves. London continued the
trip with the remaining dogs. Therefore, he reached the camp later than he
expected. The author adds: “If the missing dogs had pulled 50 miles more, I
would have been only one day late!” The task is to: a) calculate the distance
between Skagway and the camp, b) find out more detailed information about
Jack London.


• 36 squares are arranged into rectangles in as many ways as possible, using all the
squares each time. Record the length and width of each rectangle you find. Plot
these on a graph and find the relationship between the points. Which rectangle
has the smallest perimeter and why is this?
• You know how to construct regular hexagon, octagon as well as square and
equilateral triangle in a plane. Try to describe regular solids in space, draw their
nets and make paper models of them. You can compare your description with the
description in mathematical literature.


• In the game of ‘noughts and crosses’ you play with a partner on a 3 x 3 board and
to win you must have either three ‘noughts’ or three ‘crosses’ in a line. Devise
European Research in Mathematics Education I.II: Group 5 58

strategies for not losing (a) if you start the game, (b) if you are second. How
many possibly winning lines are there on a 3 x 3 board. How many winning lines
are there if you are playing on a 4 x 4 board, a 5 x 5 board etc. Could you find
how many winning lines there were on an n x n board?
• There are five numbered cubes in an open box. The box has room for just six such
cubes, so there is a vacant space (Fig. 1) which allows the cubes to be moved
around. They may not be taken out of the box in the course of the puzzle, only
sliding movements can be made. Find the least number of moves which can
transform the position in Fig. 2 into the position in Fig. 1.

1 2 3 2 4

5 4 3 Fig. 1 5 1 Fig. 2

Both projects and puzzles enable the student to:
a) develop a student’s strategic thinking and the ability to conjecture,
b) discover new relationships,
c) consolidate his/her knowledge and skills,
d) exercise hintertho imperfectly mastered skills.

We include all three types of tasks in our teaching on purpose, since it enables us to
develop mathematical thinking in our students
• of different kinds (strategic, functional, algorithmic, combinatorial, …),
• on different levels according to their individual abilities.

That are very often puzzles can be used to show pupils the enjoyment of
mathematics much more easily than any other way. They think they are playing games
and this is especially the case if the pupils have had bad experiences with mathematics
or are not confident.

3.1 Work With Projects

At present we proceed in projects mostly according to the following scheme that we
however do not consider either to be definitive, or to be exhaustive.
European Research in Mathematics Education I.II: Group 5 59


setting of

What are teaching objectives?
aims & objectives What learning objectives do you want to

Choose an appropriate topic to meet
choice of topic above aims & objectives and to be in a
context relevant to the student.

Does the topic cover
scope of topic several subjects?
mathematics only?

In school only?
time allocation
In school + in student’s own time?

Several subjects:
List areas for each subject, look for
links, develop ideas.
topic mapping Mathematics only:
What is the mathematics needed; the
background of students; the links across
different aspects of mathematics?

Gather practical equipment needed,
resource needs
reference and source books.

Open – teacher describes investigation
type of topic Closed – teacher will need to prepare
detailed brief of what has to be done
European Research in Mathematics Education I.II: Group 5 60


Teacher to discuss the subject of the topic
with students, engender enthusiasm,
preparatory stage get them to put forward ideas.
Do they want to work individually, in


Students to set out a brief of what they
intend to do and what they hope will
implementation be their outcomes.
Students experiment, teacher to act as
facilitator only.

Teacher to provide materials for
experiment with brief giving details of
preparatory stage
what and how they are expected to
experiment, after discussion of topic.


Students to work on their experiment in
implementation groups/individually as directed by the
European Research in Mathematics Education I.II: Group 5 61


Were the projects defined clearly enough
for the pupils?
Was the mathematics needed at a
suitable level for the pupils?
Did I use a suitable teaching strategy?
Teacher Did the pupils meet all the objectives I
had set?
Would I change anything when I do it
Did the students understand deeply
enough the concepts used?

Students to report back verbally to class
on their findings giving a brief
Teacher to have noted by observation the
contribution made by individuals
during the project and the final
Any outside observer.

Assessment of the student’s written work
to see whether they understood the
problem, had the ability to systematise
indirect their experiment and could ‘solve’ the
Were they able to describe what they had
found out?
European Research in Mathematics Education I.II: Group 5 62

3.2 Work With Our Tasks

In mathematics lessons our students work with project/puzzles individually or in
groups. They use different solving strategies. The main ones are (Schuh, 1968):

• Trial and error. This activity is important and useful when solving a pure puzzle.
Its disadvantage is, that it can invite the belief that solving a pure puzzle is a
matter of accident, i.e. of luck. This opinion is quite widespread, but incorrect.
Finding one solution when solving a pure puzzle, the solver cannot know
whether he/she has found all solutions or not.
• Systematic trial. This is a trial where the solver knows precisely which case has
been considered and which has not. The solver is following a logical sequence of
cases defined mainly by their mathematical knowledge.
• Division into cases. The solver divides various possibilities into groups
according to his/her own classification system. He/she works with so established
sub-cases, sub-sub-cases, … .
• Puzzle tree. The procedure of grouping various possibilities can be used in the
form of puzzle tree (example see Fig. 3).

Fig. 3: Example of a puzzle tree

4. Conclusions

Implementation of projects in actual teaching was associated for us with several
difficulties. On several occasions we had to face our colleagues’ lack of understanding
of the benefits and processes involved in teaching/using projects. But for us, it is a
fundamental fact that using projects brought such changes in mathematics
teaching/learning enabling our students to obtain positive attitudes towards
mathematics and even the weakest students becoming gradually involved in solving
European Research in Mathematics Education I.II: Group 5 63

5. References
Comenius, J.A. (1631). Didactica Magna. Reedition: Komensky, J.A. (1905). Didaktika velika.
Dedictvi Komenskeho, Praha.
Grecmanova, H. & Urbanovska, E. (1197). Projektove vyucovani a jeho vyznam v soucasne skole.
Pedagogika, XLVII, 37-45.
Bishop, A.J., Clements, K. Keitel, Ch., Kilpatrick, J., Laborde, C. (Eds.) (1996): International
Handbook of Mathematics Education. Kluwer Academic Publishers.
Koman, M. & Ticha, M. (1997). Grasping of Situations and the Development of Activity and
Cognitive Abilities. In M. Hejny, J. Novotna (Eds.): Proceedings ERCME 97. Praha,
Kubinova, M., Novotna, J., Hankova, J., Littler, G.H. & Taylor, V. (1994). Students’ Experimental
Work in Mathematics Lessons. Publications DMME, 8, Praha.
Kubinova, M. (1995). Calculative Process as a Sequence of Operations (a Particular Case). In M.
Hejny, J. Novotna (Eds.), Proceedings SEMT 95. Praha, Faculty of Education, Charles
Kubinova, M. (1997). Students’ Projects from the Point of View of Diagnostics. In M. Hejny, J.
Novotna (Eds.), Proceedings SEMT 97. Praha, Prometheus.
Kubinova, M. & Novotna, J. (1997). Students’ Independent Work in Mathematics out of School.
Mathematics Competitions, Vol. 10, No 2, 14-28.
Kubinova, M. & Novotna, J. (In print). Projects in Mathematics Education. In: Proceedings CIEAEM
50. CIEAEM, Neûchatel 1998.
Kubinova, M. & Ticha, M. (1998). On the Activating Role of Projects in the Classroom. In:
Proceedings CERME 1, WG 4. Osnabrück.
Littler, G.H. & Koman, M.: Challenging Activities for Students and Teachers. To be published 1998.
A Review of Monitoring in Mathematics, Assessment of Performance Unit (1987). DES, London.
Schuh, F. (1968). The Master Book of Mathematical Recreations. (English translation.) New York,
Dover Publ..

The paper is a part of a project GACR 406/96/1186: Mathematical education of 5 to
15-year-old students. The presentation was supported by The Ministry of Education of
Czech Republic (Project PG 98374).
European Research in Mathematics Education I.II: Group 5 64

Pearla Nesher1, Sara Hershkovitz1, Jarmila Novotna2
University of Haifa, Israel; Centre for Educational Technology, Israel;
Faculty of Education, Charles University, the Czech Republic

Abstract: Based on previous work by Novotna et al., 12 three-term multiple
comparisons problems were analyzed and field tested with teachers (in Israel) and
students (in the Czech Republic). The variables in the study were: The compared and
the reference in problems, the use of the verbal expression “more than” and “less
than”, and the underlying schemes. First analysis and the preliminary results are now
presented here.
Keywords: multiplicative compare problems.

1. The Problems

Here is a story (situation): Peter, David and Jirka together own 198 marbles: of which
David has 22 marbles, Jirka has 44 and Peter has 132. This situation is the basis for
several algebra problems. Novotna, for example used it under three different verbal
forms and studied how six-graders solve them (Novotna, 1997). Here are the forms
used by Novotna (this paper is based on the work Novotna started with Prof. Bednarz
and Prof. Janvier at CIRADE and UQAM at Montreal (Bednarz & Janvier, 1994;
Kubinova et al., 1994):

A1. Peter, David and Jirka play marbles. They have 198 marbles altogether. Peter has
6 times more marbles than David, and Jirka has twice more than David. How many
marbles does each child have?
A2. Peter, David and Jirka play marbles. They have 198 marbles altogether. Peter has
3 times more than Jirka, and Jirka has twice more than David. How many marbles does
each child have?
A3. Peter, David and Jirka play marbles. They have 198 marbles altogether. Peter has
6 times more than David, and 3 times more than Jirka. How many marbles does each
child have?
European Research in Mathematics Education I.II: Group 5 65

In her study Novotna found that though the same situation is described, the students
performances are different according to the form used to present these three problems.
Our analysis led us to formulate another nine possibilities for the same underlying
situation. The next three problems are similar to Novotna’s problems yet they describe
the relationship between the boys, this time in terms of a “less than” relation. The other
six problem mention both relations “more than” and “less than” in the same problems.
B1. Peter, David and Jirka play marbles. They have 198 marbles altogether. David has
6 times less than Peter, and he has twice less marbles than Jirka. How many marbles
does each child have?
B2. Peter, David and Jirka play marbles. They have 198 marbles altogether. David has
twice less than Jirka and Jirka has 3 times less than Peter. How many marbles does
each child have?
B3. Peter, David and Jirka play marbles. They have 198 marbles altogether. David has
6 times less than Peter, and Jirka has 3 times less than Peter. How many marbles does
each child have?
C1. Peter, David and Jirka play marbles. They have 198 marbles altogether. Peter has
6 times more than David, and David has twice less than Jirka. How many marbles does
each child have?
C2. Peter, David and Jirka play marbles. They have 198 marbles altogether. Peter has
3 times more than Jirka, and David has twice less than Jirka. How many marbles does
each child have?
C3. Peter, David and Jirka play marbles. They have 198 marbles altogether. David has
6 times less than Peter, and Peter has 3 times more marbles than Jirka. How many
marbles does each child have?

E1. Peter, David and Jirka play marbles. They have 198 marbles altogether. Jirka has
twice more marbles than David and David has 6 times less than Peter. How many
marbles does each child have?
E2. Peter, David and Jirka play marbles. They have 198 marbles altogether. Jirka has
3 times less marbles than Peter, and twice more marbles than David. How many
marbles does each child have?
E3. Peter, David and Jirka play marbles. They have 198 marbles altogether. Jirka has
3 times less marbles than Peter, and Peter has 6 times more than David. How many
marbles does each child have?
European Research in Mathematics Education I.II: Group 5 66

2. Analysis of the Problems

All the above problems are of the type known to be “multiplicative compare problems”.
As already mentioned (Nesher, Greeno, & Riley, 1982) the compare problems are
analyzed in terms of a two-place relation R(a,b) in which ‘a’ is the compared quantity,
and ‘b’ is the reference quantity. The comparison relationship can be described by
either “more than” or “less than”. One should note that the relation “more(a,b)”
describes the same relationship as “less(b,a)”, in which the arguments changed role, the
compared becoming the reference and the reference becoming the compared quantity.
To say “David has less than Peter” is the same as saying “Peter has more than David”.
The comparison relation is asymmetrical and in most cases we have the choice as to the
wording to describe linguistically the situation; that means we may decide what will be
for us the compared quantity, and what will serve as the reference. The compared
quantity will then be the subject in the sentence that describes the situation, and the
reference will be part of the predicate.

From this point of view, each of the above 12 problems, involves three comparison
(a) The non-ordered (the non-ordered relation between David and Peter, is going to be
an ordered relation once we use the verbal description “more” or “less”)
comparison between David and Peter R(D,P), or R(P,D).
(b) The non-ordered comparison between David and Jirka R(D,J), or R(J,D),
(c) The non-ordered comparison between Peter and Jirka R(P,J), or R(J,P),

where P, J and D stand for Peter, Jirka and David, respectively.

In the above problems (A1-D3) the total amount of marbles is given. This sows that
in order to present a problem that can be solved, only two relationships of comparison
are needed. There are altogether three different combinations of two relations.
Therefore, in all the above problems we will find one of the following combinations of
the basic comparisons:

I. The problem mentions the non-ordered comparison between R(D,P) and R(D,J).
(Problems: A1, B1, C1, E1)
European Research in Mathematics Education I.II: Group 5 67

II. The problem mentions the comparison between R(D,J) and R(P,J).
(Problems: A2, B2, C2, E2)
III.The problem mentions the comparison between R(D,P) and R(P,J).
(Problems: A3, B3, C3, E3).

As explained before, each of the comparison relationship is asymmetrical. Then,
with the use of “more than” and “less than” for each of the basic relations (a-c), we
arrive at 12 different formulations of word problems for the same underlying structure,
for the same situation described in the beginning of this paper.

3. Variables Involved in the Study

3.1 The Compared and the Reference

As we have already mentioned, each problem consists of two explicit compare relations
(see, I - III). Yet, the direction of the description and the wording used in the
comparison are different and they will be distinguished. For example, in A1 the relation
“more than” is employed in both relations. Thus, the situation dictates that David will
be the reference and Peter and Jirka the compared. Thus, in this problem two (Peter and
Jirka) are the compared to one reference (David). In problem B2, although it is again a
type-II problem as explained before, here, because of employing the relation “less
than” it is David, the compared and Jirka and Peter are the reference.

The choice of the “compared” or the “reference” dictates which will appear as the
subject of the sentence (in the word problem) and which will be part of the predicate.
Table I describes the compare-reference situation in each problem the head of the arrow
points to the reference.

From this point of view we could distinguish three types of problems, those with one
compared and two references (we mark it “1 to 2”) (problems: A3, B1, E2), those with
two compared and one reference (“2 to 1”) (problems A1, B3, B2), and those with 2
compared and 2 references (“2 to 2”) (problems: A2, B2, C1, C3, E1, E3).
European Research in Mathematics Education I.II: Group 5 68

A1 2:1 M C1 2:2 ML
A2 2:2 M C2 2:1 ML
A3 1:2 M C3 2:2 ML
B1 1:2 L E1 2:2 ML
B2 2:2 L E2 1:2 ML
B3 2:1 L E3 2:2 ML

Tab. I: Analysis of the problems

From the syntactic point of view, the “2 to 2” format problem text are a conjunction
of two full sentences. The predicate of the first sentence becomes the subject of the
second sentence. These problems seem to be the easiest ones to grasp from the point of
view of the flaw of information. Problems of the “2 to 1” format, are a conjunction of
predicates, and as those of the “1 to 2” format, they make an anaphoric mention of the
same subject which plays a role in the two comparisons.

3.2 “More Than” or “Less Than”

Several studies suggest that the word “more” is more easily comprehended than the
word “less” (Donaldson & Balfour, 1968; Nesher & Teubal, 1975). In our case
regarding three of the problems (A1, A2, A3), we have use the word “more”. Then in
three problems (B1, B2, B3) we have used the word “less” and as for problem (C1-E3)
we have used both concepts together.

At this point, we need to add a special comment. In English both the words “more”
and “less” may describe an additive comparison (such as “Ron has 5 stamps more than
Rene”, or Rene has 5 stamps less than Ron). For multiplicative comparison, however,
the expression used is : “Ron has 5 times stamps as Rene”. The words “more” and
“less” do not usually appear in a multiplicative comparison. But, in our languages
European Research in Mathematics Education I.II: Group 5 69

(Hebrew and Czech) the words “more” and “less” are used for the multiplicative
compare relation as well as for the additive compare relation (Nesher, 1988).

3.3 The Underlying Scheme

In earlier studies by (Nesher & Hershkovitz, 1994) (Hershkovitz & Nesher, 1996) we
have demonstrated the role of schemes in explaining the variance in students
performance. The schemes were of three types:

(1) Hierarchical,
(2) Sharing parts, and
(3) Sharing whole.

Figure 1 presents these three schemes.

Shared Part Hierarchical Shared Whole

Fig. 1: The three schemes

An analysis of Novotna’s problems (E3, A2 and A3) has shown that each one of
them belongs to a different scheme type. All share the fact that the quantity the children
have altogether are 198 and we will not include it now in our presentation which is
distinct for each problem.

Along this analysis we can map the set of 12 problems as follows:
The Shared-part scheme: Problems: A1, B1, C1, E1.
The Hierarchical scheme: Problems A2, B2, C2, E2.
The shared-whole scheme: Problems A3, B3, C3, E3.
European Research in Mathematics Education I.II: Group 5 70

To conclude:
Following Novotna studies the set of 12 problems, lends itself to a study of the role of
each of the above variables. The 12 problems were given to experienced teachers as
well as to students.

3.4 The Variables of the Study:

1) The scheme: Shared-Part (SP), Hierarchical (H), or Shared-Whole (SW).
2) The use of the relation “more than (M); “less than”(L), both (ML).
3) The syntactic combination of “Compared” and the “Reference” in the compound
problems (“1 to 2”, “2 to 1” and “2 to 2”).

4. The Experiment

The set of twelve problems was given to about 100 teachers in an in-service workshop
in Israel and to students in the Czech Republic. The teachers were all experienced in
teaching mathematics in primary schools. The twelve problems were arranged in three
separate forms by type, each containing only four problems from different classes.
Each teacher solved only one form. The forms were distributed at random.

Each problem was solved by about 35 teachers. It took about 40 minutes for to
complete the task.

5. Findings

Table II presents the raw data in percentage obtained for each problem by the teachers.

Problem # A1 A2 A3 B1 B2 B3 C1 C2 C3 E1 E2 E3
%s success 100 86 74 67 78 92 86 81 76 76 58 80

Tab. II: Percentage of Success for Each Problem
European Research in Mathematics Education I.II: Group 5 71

More statistical elaboration concerning the role of each variable will be presented at
the conference.

5.1 Strategies

The following strategies were observed:
I. Algebraic methods
II. Arithmetic methods

Within the algebraic methods, teachers used one variable but they each assigned
a different quantity to the reference in the equations.

The references were:
I.1: Choosing the smallest quantity (D) consistently to be the reference in any
problem they solved.
I.2: Choosing the smallest quantity (D) to be the reference when the comparisons
were given by the word “more” and choosing the largest quantity (P) to be the
reference when the comparisons were given by the word “less”.
I.3: Czech students used equations with two unknowns.
I.4: We could not find consistency.
II. The arithmetic strategy, consisted of:
II.1. Using a ratio strategy.
II.2. Division by the number of participants (used by Czech students).

Results from problem A3 solved in the Czech Republic in 1994, are presented in
table III.
European Research in Mathematics Education I.II: Group 5 72

Grade 4th 5th 6th 7th

No. of solver 50 68 241 20

Successful 1 9 30 12

Unsuccessful 49 59 211 8

Arithmetical (Ar1) 9 17 51 0

Arithmetical (Ar2) 7 6 62 1

Algebraic (1 equation) 4 15 3 17

Algebraic (2 equations) 1 1 0 1

Algebraic (3 equations) 2 2 2 0

Non-scholar 2 1 0 0

Not solved by 25 26 123 2

Tab. III: Results from the Czech Republic

Ar1:Calculated using 1 part (David) (198:9=22 ; 22X6=132 ; 22X2=44)
Ar2:Calculated using number of participants (198:3= 66) then trying to satisfy the
given relationship.

Note: Unsuccessful solutions were mostly probably influenced by the following
reason: There was another problem in the set, where two participants were present and
where the strategy was successful (The total fee which Mr. Novak’s daughters, Pavla
and Marie, received was 181 CZK. Marie had 37 CZK more then Pavla. How many
crowns did each daughter receive?)

Algebraic, 1 equation (David x, Peter 6x, Jirka 6x : 2 = 3x) and modifications
Algebraic, 2 equations (David x, Peter 6x, Jirka y, 6x = 3y) and modifications
Algebraic, 3 equations (David x, Peter z, Jirka y, z = 3y, z = 6x) and modifications

Non-scholar: use of “approximation”: the student estimated David’s amount,
calculated Jirka’s and Peter’s amount following the assigned relationship, then he/she
added the three amounts and compared the sum with the assigned sum of marbles. In
case of error in the sum, he/she restarted with another starting number.
European Research in Mathematics Education I.II: Group 5 73

Note: This strategy we found rarely used, but after using it once, the student used it
systematically to solve all the assigned problems in the set.

Preliminary analysis indicates that some of the analyzed before variables are
dependent. For example, the choice of reference is dependent on the problem’s
formulation, whether it uses the words “more” or “less”. It seems that the compared and
the reference in the problem were of greatest importance influencing the solver. About
90% of the teachers used algebraic strategies and only about 10% used arithmetic
strategies. The most frequent algebraic strategy reported was the one where David was
the reference, even when the text used others names as the reference.

6. References
Bednarz, N. & Janvier, B. (1994). The Emergency and Development of Algebraic in a Problem
Solving Context: An Analysis of Problem. In Proceeding Int. Group for Psychology in
Mathematics Education. Lisbon.
Donaldson, M., & Balfour, G. (1968). Less is More? A Study of Language Comprehension in
Children. British Journal of Psychology, 59 (4), 463-471.
Hershkovitz, S., & Nesher, P. (1996). The Role of Schemes in Designing Computerized
Environments. Educational Studies in Mathematics, 30, 339-366.
Kubinova, M., Novotna, J. Bednarz, N., Janvier, B. & Totohasina, A. (1994). Strategies Used by
Students when Solving Word Problems. Publications DMME, 9, Praha.
Nesher, P. (1988). Multiplicative School Word Problems: Theoretical Approaches and Empirical
Findings. In J. Hiebert & M. Behr (Eds.), Number Concepts and Operations in the Middle
Grades (pp. 19-41). NJ: Lawrence Erlbaum Association.
Nesher, P., Greeno, J. J., & Riley, M. S. (1982). The Development of Semantic Categories for
Addition and Subtraction. Educational Studies in Mathematics, 13, 373-394.
Nesher, P., & Hershkovitz, S. (1994). The Role of Schemes in Two-step Problems: Analysis and
Research Findings. Educational Studies in Mathematics, 26, 1-23.
Nesher, P., & Teubal, E. (1975). Verbal Cues as an Interfering Factor in Verbal Problem Solving.
Educational Studies in Mathematics, 6, 41 - 51.
Novotna, J. (1997). Using Geometrical Models and Interviews as Diagnostic Tools to Determine
Students’ Misunderstanding in Mathematics. In: Proceedings SEMT 97. Praha, Prometheus.
Novotna, J. (1997). Geometrical Models in Solving Word Problems that Include the Division of the
European Research in Mathematics Education I.II: Group 5 74

Whole into Parts (Theory and Practice). In: Proceedings Interakcja teorii i praktyki nauczania
matematyki w szkole podstawowej i sredniej VSP Rzeszow.
Novotna, J. (1997). Phenomena Discovered in the Process of Solving Word Problems. In:
Proceedings ERCME 97. Praha, Prometheus.
Novotna, J. (1997). Vizualni modelovani jevu “rovnice” In: Sbornik K aktualnim otazkam
matematicke pripravy ucitelu 1. stupne ZS (OS) na Pedagogickych fakultach v CR a SR.
Olomouc, Pedagogicka fakulta UP Olomouc. (In Czech).


The paper is a part of a project Charles University Grant Agency: Analyza kognitivnich
procesu zaku pri reseni slovnich uloh. The presentation was supported by The Ministry
of Education of Czech Republic (Project PG 98374).
European Research in Mathematics Education I.II: Group 5 75

Tatyana Oleinik
Kharkov State Pedagogical University named after G.S.Skovoroda,
pr. Lenina, 19-a, kv.25, Kharkov - 310166, Ukraine or

Abstract: This paper represents the results of special courses given to undergraduate
teacher students of «mathematics-computer science» speciality. A general idea that
integrated these courses, consist in the training of future teachers for understanding
the possibilities and limits of the use of technologies (systems like DERIVE and
Cabri-geometre) for support of pedagogical control of learners’ cognitive activity. It is
very important to develop teacher students’ abilities for their researches on why and
how they should use technologies particularly like microworlds and computer algebra
systems. It gave an opportunity to encourage teacher students’ interest in teaching
algorithmic and half-algorithmic problems as well as heuristic problems and therefore
heuristic methods of operations that intrinsic of creative activities.
Keywords: cognitive approach, mathematical thinking and learning, teacher studies.

1. Introduction

Modern education is oriented towards making students’ creative abilities more active.
This necessitates psychological and pedagogical researches devoted to the
development of teacher students’ creative capacities. To solve this complex problem, it
is not sufficient to give students firm knowledge in theory of pedagogy, psychology,
mathematics, etc., but it is necessary to mould and develop their creative features
essential for independent researches.

This paper represents the results of special courses given to undergraduate teacher
students of «mathematics-computer science» speciality. A general idea that integrated
these courses, consist in the training of future teachers for understanding the
possibilities and limits of the use of technologies (systems like DERIVE and
Cabri-geometre) for learners’ cognitive activity. It is very important to develop teacher
European Research in Mathematics Education I.II: Group 5 76

students’ abilities for their researches on why and how they should use technologies
particularly like microworlds and computer algebra systems (CAS).

Since 1992 a special course “The use of mathematical packages for students
explorations in Mathematical Analysis” has been worked out for undergraduate teacher
students in their future classes for discovering and forming mathematical concepts.
There is no doubt that teaching based on computer explorations contributes to the
growth cognitive activity, but in non-computer variant it was out limits for the majority
of teacher students.

But with the advent of improvement technologies they could feel its advantages and
realize its essence. It gave an opportunity to encourage teacher students’ interest in
problems of pedagogical control of pupils’ cognitive activity. It is necessary to modify
programs and standards of the mathematics teachers training that concern using of
technologies in their future pedagogical activity since modern software can promote
modification of view on the essence of a pedagogical activity. In expectation of
outcomes of long-term pedagogical investigations the inaction is short-sighted and
inefficient therefore future teachers have to get preparation for their own pedagogical

2. Background

Our previous researches (Rakov & Oleinik et. al. 1994, Oleinik et. al. 1996) has shown
that it is insufficient for a student to have good knowledge, skills and habits. It is
necessary to develop a person’s psychological peculiarities essential for a certain
activity and contributing to its successful fulfillment. They manifest themselves in a
person’s intellectual qualities system: the ability generate new ideas and put forward
problems independently, flexibility and originality of thinking (the ability to perceive a
well-known problem in new context), the ability to transpose knowledge and skills in a
new situation (towards a new problem), etc.

Our approach is based on mathematical thinking and learning theory of Schwank
(1986, 1993) about individual preferences in mental structure, namely predicative
versus functional cognitive structures. The teaching process has to reflect these results.
European Research in Mathematics Education I.II: Group 5 77

The textbooks and results of researches under guidance Cohors-Fresenborg
(Cohors-Fresenborg 1993, Cohors-Fresenborg et. al. 1991 a, b) are a good proof of
success of this theory and an important contribution to it. These are very significant for
understanding the process of mathematical concept formation. Main ideas and results
have formed the foundation of one of our special courses and attracted students

It goes without saying that theoretical framework for our research was constructivist
understanding of teaching and learning. We tried to orient teacher students’ main
efforts towards researching problems connected with the moulding (formation) of
cognitive activity methods on the basis of already acquired skills and methods. Though
we know that this process is closely connected both with theoretical system of
knowledge and with creative thinking.

The first component of students’ approach is based on a theory of visual thinking
and its development (Arnchame 1981, Luria 1981). The most important stage of visual
thinking is a stage of mental framing hypothesis about possible ways of solving
problems with analyzing and forecasting their possible results.

The second component is a theory of mathematical creativity development
(Krutetzkii 1968) and paradigm of the open-ended approach in mathematics teaching (
Nohda 1991, Pehkonen 1995) for the development general competency of scientific
style of activity (abstraction, generalization, specialization etc.).

3. The Experiments and Method

Altogether about 70 students were involved in our experiments during 2 years. Every
year it was one group of fourth year students and one group of fifth year students. It is
natural that during the second year of experiment it was the experimental group that
was chosen (the one that had already taken part in the experiment a year ago).
European Research in Mathematics Education I.II: Group 5 78

Our hypothesis was that the special courses offered form teacher students’ creative
features, contribute to successful fulfillment of their own pedagogical researches on
cognitive activity. Besides, we thought that using microworld Dynamic Mazes allow
to show different cognitive ways of thinking and define which students have
difficulties for predicative way of thinking or for functional one. We thought that
study of these ideas would greatly increase the effectiveness of teacher students’

Every special course consisted of 30 hours during the spring academic term. During the
first year the following special courses were offered:
1. Main ideas and results of curriculum project «Integrating Algorithmic and
Axiomatic Ways of Thinking in Mathematics Lessons for grade 7 and 8» using
microworld Dynamic Mazes (for the students experimental group of 4th course).
2. Computer Explorations in Math Courses (Plane Geometry, Algebra and Math
Analysis) (for the students of 5th course).
During the second year the following special courses were offered:
1. Using Derive system for Solving Problems with Economic Contents (first part) and
for Trigonometric Equation Solving (second part) (for the students of 4th course).
2. Computer Explorations in Plane Geometry (first part) Solving Problems with
Economic Contents Using Derive system (second part) (for the students
experimental group of 5th course).

Thus, one experimental group during two years was given two special courses first
of which was connected with the study on research of individual differences in students
methods of mental analysis using microworld Dynamic Mazes.

Students performed their course and diploma papers in a constant cooperation and
collaboration with their research advisor. Students chose their themes independently
out of the proposed list, they also chose the complexity and depth of their researches
during its process.

The process of teaching and learning mathematics is divided into three range levels
(though all of them are connected the first one is more complicated):
• development of creativity as research thinking;
European Research in Mathematics Education I.II: Group 5 79

• acquisition of cognitive activity methods;
• acquisition of pure mathematical knowledge and skills.

According to the above mentioned, we have distinguished three range down levels
of students’ researches on each of them pedagogical problems imminent to each type of
activities were solved: 1) the creativity of non-standard algorithms, independent
making of problems, generalization of facts, phenomena, laws, strategies; 2) the
activities based on an already known algorithm but with new contents; 3) the
reproductive activity (reproducing according to the pattern). The successful solution of
these problems was connected with the development of respective intellectual skills.

As our research has shown, the students from the experimental group moved to the
first level quicker and easier than others. Besides their diploma papers were
characterized by original and independent style, though themes chosen were complex
enough. For example: formation pupils’ skills and habits of solving trigonometric
equations or solving problems of real life, computer explorations in studying surfaces
of second order or plane curves.

Besides, the majority of the experimental group mastered the second level. In two
other groups there were no results of the kind (though these two groups were different
as far as their marks were concerned).

3.1 Some Remarks About Essence of Students’ Researches

Computer explorations are very important for the process of studying (cognitive
activity) as well as computing experiment for the process of scientific cognition. In the
first case, students study subjectively unknown facts, in the second - scientists
investigate objectively unknown facts of nature. Therefore teacher students try to
develop pupils’ exploring ability and to guide (to help) in organizing this process
according to some steps: posing out problems; making students’ explorations based on
computing experiments; framing hypothesis about the way of solving; proving the
hypothesis or creating the counterexample to it.
European Research in Mathematics Education I.II: Group 5 80

Computer explorations become more effective by Derive and Cabri-geometre
facilitating all these steps (graphing and geometric transformations, accurate counting
and equations solving, simplifying which produce equivalent expressions and choosing
direction of simplifying, etc.). But students have to know many new things, for
example: how to prove theorems of well-based hypothesis that close connected with
constructing counterexamples (which are capable to refuse wrong hypothesis) or how
to choose a compromise between calculation speed and results accuracy, common
methods and solution errors. Besides for forming students ability of posing out
problems a fruitful approach is a system of problems arranged in series so that the
experience of solving the previous problem helps to pose the next one. We suggest
(Oleinik et. al.1996) organizing the explorations in four levels for development of
exploring pupils’ competencies, which provides a success in development of
mathematical creativity.

In plane geometry we propose some series of arranging problems which help in
posing out problem (and forming hypothesis), for example:
• theorems on the equality (or congruence conditions) of the triangles and
theorems on the similarity of the triangles (as for any triangles as for right
triangles), properties of sides and angles in triangle;
• theorems on the equality (or congruence conditions) of the angles formed by two
parallel lines and intersecting line (internal and external crosswise lying angles,
internal and external corresponding angles etc.), theorems on the sum of the
angles of triangle; value of external angle of triangle;
• area of right triangle, area of any triangle, Pythagorean theorem, properties of
perpendicular, inclined line and projection.

For the acquisition of methods for solving real world problems (applied
mathematics) was chosen economical application because many young people are very
interested in market economy (ability to solve economical problems is very important
in contemporary life). This is the main reason why we have used these topics in the
beginning of mathematical analysis teaching and learning. We hope that this approach
will provide more than ordinary efficiency of understanding mathematics. Besides for
Derive allow rearranging the work time in learning, in particular, we create time to
teach new topic “modeling-translating-interpreting”.
European Research in Mathematics Education I.II: Group 5 81

Computer explorations can be successfully used in solving real life problems as
well. Therefore it is important for forming suitable students ability in applied
mathematics courses that requires four steps, as usual: choosing a suitable model;
translating the real world problem into the mathematical problem; calculating the
model solution by applying known methods; interpreting the model solution into a real
world solution. For teacher students’ researches in this direction we propose to use the
method of constructing and reconstructing of problems, to compare the obtaining
results from different ways of problem solving, to generalize solving method of this
type of problems etc.

For example one economical problem which use equations of second order curves
and two reconstructing questions (open-approach method): «1. Investigate (find) the
solution of the previous problem when the price of 1 unit of production of the
enterprises A and B are change (for example, to accordingly 200 and 225 units of
money). 2. Investigate (find) the solution of the previous problem when the transport
cost for 1 km is change (for example, from the enterprise A is 2 times less than B).»

The research on possibilities of a system DERIVE concerning the management of
trigonometric transformations has shown that they are unique for teaching traditional
school methods of their solution (factorization, reduction to a quadratic equation of one
functions or homogeneous etc.) as well as more complicated methods.

A teacher’s important task is to teach his pupil to see the thing that is instilled in
images, i.e. to analyze visual information. It is the discovery of certain fragments and
the identification of similar ones (either in form on meaning) that take place first. But
the working out of problem solving plan is the most important stage. Derive is very
useful in this process as well as in generalization methods of equation solving
(discovery of algorithms).

For teacher students’ research of in this direction we propose to reconstruct a pupil’s
possible way of thinking (and to form a pupil’s visual thinking) while analyzing the
following problem: «To solve the equation 3sin2x - 2sin2x +5cos2x = 2.
1. The equation includes trigonometric functions that are why it is a trigonometric
2. The equation includes different trigonometric functions with different arguments.
European Research in Mathematics Education I.II: Group 5 82

3. All the summands of the both sides can be represented as a function of one
4. All the summands have a similar degree and we can divide by cos2x for obtaining an
equation of one trigonometric function.
5. We know two types of equations simplest ƒ(x)=a and more complicated
aƒ2(x) + bƒ(x) + c = 0.
6. We obtain quadric equation of one trigonometric function.»

The next problem of research is geometric applications of complex numbers (CN)
unfortunately which was forgot in a school curricular. Though the method CN allows
solving plane geometric problems by the elementary calculations (using the known
formulas) which immediately follow from the problem condition. Therefore students’
interest to a method CN with using DERIVE is connected to the greater simplicity of its
application in a comparison with traditional coordinate and vector methods, which
demand considerable quickness of pupils. Even few simplest statements (that are
follow from the geometric interpretation of complex numbers) allow to solve rather
useful problems on the proof of properties of triangles and tetragons. Besides they
allow to prove the known classical theorems of elementary geometry.

4. Some Preliminary Results

The experience realizations these special courses shows, that the use DERIVE and
Cabri-geometre support in teaching algorithmic and half-algorithmic problems as well
as heuristic problems and therefore heuristic methods of operations that intrinsic of
creative activities (abstraction, generalization, specialization etc.).

The evaluation of experiment data allows us to formulate the following results.

1. Students’ good knowledge does not guarantee their successful creative activity.
2. Technologies like microworlds and computer algebra system contribute to growing
teacher students’ interest in psychological researches on cognitive activity.
European Research in Mathematics Education I.II: Group 5 83

3. Studying modern teaching methods (for example, the levels and methods of guiding
explorations, heuristic methods etc.), the students develop new problems by
themselves and investigate new methodical ideas with using technology.
4. Studying essence of the theory of individual preferences in mental structure by
computer microworld increases the effectiveness of students’ researches and shows
great students’ interest to the development of practical recommendations of this
5. The considered approaches provide a success in teacher students’ researches on
cognitive activity (“open-end” problems in algebra and geometry; complex
numbers and their geometric applications; elements of linear programming; moving
and rotation objects on the plane (tangram) etc.)

5. References
Arnchame, R. (1981): Visual Thinking. Psychology of Thinking. Moscow.
Cohors-Fresenborg, E. (1993): Registermachine as a Mental Model for Understanding Computer
Programming. Berlin.
Cohors-Fresenborg, E. & Kaune, C. & Griep, M. (1991 a, b): Einführung in die Computerwelt mit
Registermaschinen, Textbuch fur Schüler, Handbuch fur Lehrer. Osnabrück.
Luria, A. (1981): Mnemonic mind. Psychology of Thinking. Moscow.
Krutetzkii, V. (1968): Mathematical Abilities Psychology of Schoolchildren. Moscow.
Nohda, N. (1991): Paradigm of the “open-approach” method in mathematics teaching: Focus on
mathematical problem solving. In: International Reviews on Mathematical Education. 23 (2).
Oleinik, T. & Krikun, V. (1996): School Geometry Explorations with Using TRAGECAL Package.
Kharkov (in Russian)
Pehkonen, E. (1995): Introduction: Use of Open-Ended Problems. In: International Reviews on
Mathematical Education. 27 (2).
Rakov, S. & Oleinik, T. etc. (1994): Discovery Calculus with Technology. Kharkov. (in Russian)
Schwank, I. (1986): Cognitive Structures of Algorithmic Thinking. In: Proceedings of the 10th
Conference for the Psychology of Mathematics Education, 195-200. London.
Schwank, I. (1993): On the Analysis of Cognitive Structures of Algorithmic Thinking. Journal of
Mathematical Behavior.
European Research in Mathematics Education I.II: Group 5 84

Inge Schwank
Institute for Cognitive Mathematics, Department of Mathematics and Computer Science,
University of Osnabrueck, D-49069 Osnabrueck

Abstract: Predicative thinking is thinking in terms of relations and judgments;
functional thinking is thinking in terms of available actions and achievable effects.
Depending on the way of thinking the orientation in the world, the type of sources for
getting insight are not the same. E.g. it should be visible in different eye movements. In
addition to our qualitative experiments, recently we started to run a study based on
EEG-methods while students were solving logical pattern fitting tasks. The EEG
complexity during predicative thinking decreased in comparison to functional thinking
and mental relaxation, with this reduction being most pronounced over the parietal and
right cortex. A reduction in dimensional complexity during functional thinking which
was concentrated over the left central cortex, although significant, was less clear.
Keywords: cognition, EEG, eye movement.

1. Introduction

Concerning cognitive models as sources for intelligent behaviour like planning or
learning different approaches have been studied during the years. There is broad
agreement - since the time of Plato - that language and - as a newer approach in our
century - visualization is a tool of thinking. Together with the cognitive turn as
movement against behaviourism lots of statements indicating that thinking equals
language were set.

There are good reasons to accept another cognitive base of elaborated thinking,
which is related in a specific sense to motoric-guided cognitive representations. At an
interdisciplinary conference on “Thinking and Speaking” the mathematician van der
Waerden (1954) pointed out, that thinking in motoric terms is on top level in creative
mathematical work: it might be that someone is a visualizer and therefore has more
European Research in Mathematics Education I.II: Group 5 85

benefit in thinking in images, but for sure the words which are used to label the
mathematical concepts only play a minor role. This seems to be in contrast e.g. to what
is used of the Piagetian theory of levels of cognitive development, where motoric
representations are on the bottom of the hierarchy and it is the formal representations
which are on the top. In our theory we use the term functional thinking for motoric
thinking in the sense of van der Waerden (which includes only such motoric actions
useful for productions) and we contrast it with a kind of thinking that doesn’t care so
much on dynamics but on static structures and the embedded complex relationships,
this latter we call predicative thinking. Up to now there is not much clearness about the
relationship of functional/predicative thinking and all kinds of visualization and
imagery. It seems to be clear that in case of predicative thinking good language
knowledge is useful, because it is of much advantage in creating an adequate structure
in a problem solving situation to use word-labels to build up this specific structure.

Beyond the fact that the functional way of thinking plays an important role in
mathematical thinking its range of appliance is even much broader. Bateson (19804,
120-121) presented a nice example concerning the difference we are interested in:

»We do not notice that the concept “switch” is of a quite different order from the
concepts “stone,” “table,” and the like. Closer examination shows, that the switch,
considered as a part of an electric circuit, does not exist when it is in the on position.
From the point of view of the circuit, it is not different from the conducting wire which
leads to it and the wire which leads away from it. It is merely “more conductor.”
Conversely, but similarly, when the switch is off, it does not exist from the point of
view of the circuit. It is nothing, a gap between two conductors which, themselves exist
only as conductors when the switch is on. In other words, the switch is not except at the
moment of its change of setting, and the concept “switch” has thus a special relation to
time. It is related to the notion “change” rather to the notion “object”.«

In some studies (e.g. Schwank 1993a, Schwank 1994) on cognitive ways of learning
basic concepts of computer science we used within a set of bricks (Dynamic Mazes, a mechanical switch (Fig.
1) for representing a tool for case distinction. For instance, by means of this switch
organisational problems like automatical bottle-selling procedures can be solved. The
key point is to plan actions and thereby to anticipate their influence on later occasions,
European Research in Mathematics Education I.II: Group 5 86

which is a typical functional requirement. Predicted difficulties of part of the students
based on their individual problems dealing with functional concepts arose (e.g.
Schwank 1994).

to switch to
the state »left«

switch in state »right«

to switch to
the state »right«

switch in state »left«

Fig. 1: Switch of the Dynamic Mazes
(Test it.)

In the following we first give a short overview of our theory and than present some
examples of short logical tasks, which we are using at present to check abilities and
preferences of subjects in using a functional or a predicative cognitive structure.
European Research in Mathematics Education I.II: Group 5 87

given in a specific microworld

ò ò ò ò ò ò

Cognitive grasping by means of
the preferred activated cognitive structure

ê ê
Predicative Functional

Grasping Grasping
structures / concepts structures / concepts
by means of by means of
predicates / relations functions / operations
between different on different
(mathematical) (mathematical)
objects objects
Thinking in Relations Thinking in Effects

ê ê
Development of an Development of an
internal conceptual internal conceptual
representation for representation for
static grasping dynamic grasping

Fig. 2: Predicative versus functional cognitive organisation (cf. Schwank 1995)
European Research in Mathematics Education I.II: Group 5 88

2. Predicative versus Functional Cognitive Structures

We distinguish between static and dynamic mental modelling as a characteristic of the
individual cognitive structure in terms of predicative versus functional thinking (Fig.
2). Predicative thinking emphasizes the preference of thinking in terms of relations and
judgments; functional thinking emphasizes the preference of thinking in terms of
courses/effects and modes of action (cf. Schwank 1993a, 1996). For an overview of the
experimental testing of the theory see Schwank (1995, 108-115). Research has also
shown that it is quite rare to find female subjects who behave in a functional way (see
also Schwank, 1994) or that functional and predicative thinking occurs as well in
Indonesia (Marpaung 1986) and China (Xu 1994).

The given diagram (Fig. 2) has to be read spirally in chronological order. The black
arrows describe circles in order to consider that the internal tools of the conceptual
representation influence that which will be grasped cognitively. In consequence the
further development of the internal conceptual representations is interfered. The
observed differences in behaviour partly are explained in such a way that both kinds of
cognitive structures are not applied equally which results in a different development of
a more static or a more dynamic internal conceptual representation.

The category of individual cognitive structures has to be separated from the
category of individual cognitive strategies. We distinguish between a conceptual,
top-down organising, and a sequential, more interactive approach (Cohors-Fresenborg
& Schwank, 1996). Predicative / functional refers to the tools of thinking, conceptual /
sequential refers to the global organisation of the problem-solving process. Concerning
links to other cognitive theories such as declarative/procedural see e.g. Schwank

3. Examples

For a predicatively structured person the central point of his or her analysis concerning
a complex situation is to break it down into different conceptual pieces and to invent a
logical structure which describes the network of the relations between these pieces. For
European Research in Mathematics Education I.II: Group 5 89

a functionally structured person the central point is to arrange the going through the
production as a complex process in which different strengths control, determine or
promote each other. For the former the mental model describes the logical structure, for
the latter it describes the organisation of work flow in time.

To show the benefit of our cognitive theory for e. g. a cognitive approach in business
reengineering (Cohors-Fresenborg & Schwank 1997) or a cognitive science approach
in computer programming (Schwank 1993 a,b) we have designed different studies
which are run with single subjects using different settings: fitting figures in matrices
(QuaDiPF), organising processes in a microworld (OPM). QuaDiPF (Schwank, 1998)
is a qualitative diagnostic-instrument to determine the preferred cognitive structure,
predicative versus functional. In OPM those tested have to solve a sequence of
organisational problems with the specific microworld Dynamic Mazes (cf. Cohors-
Fresenborg, 1978, This is
the mechanical realisation of a mathematical idea of automata which is equivalent to
the Turingmachine. We know from our studies that this setting in the beginning
supports the functionally structured subjects. While solving the more complex
problems a predicative cognitive structure is more successful. Here we concentrate on
QuaDiPF because the setting is much simpler and not as time consuming as OPM.
Finally this newer analysing tool is even usable in EEG-measurement-environments.

3.1 Fitting Figures in Matrices: QuaDiPF

We use tasks such as those in common intelligence tests (e. g. Raven, 1965) to find a
missing figure, which fits suitably into a set of 8 given figures arranged in a matrix. In a
clinical interview each subject has to invent and draw the missing figure in the matrix
(instead of selecting it from a given set as usual). The subject has to argue why he or she
drew this very figure. The analysis of the videotapes shows that a predicative and a
functional way of mentally modelling the task exist. In a predicative mental model the
subject uses predicative tools, e.g. looking for properties, inventing general laws. So, in
the given example (Fig. 3a) the subject tries to structure the image. Each figure consists
of three objects: a star, a point and a circle. The triangle is the same in each figure. In
each row the point is at the same place. In each column the circle is at the same place. In
European Research in Mathematics Education I.II: Group 5 90

a functional mental model the subject uses
functional tools, e.g. invents a process which
produces the last element in a row or
column. In each row the circle moves
around, and in each column the point moves
around. The object around which the
movement takes place does not change. In
both ways of dealing with the problem the
result is the identical.
Fig. 3a: QaDiPF-Example
(Schwank, 1998) Besides tasks such as 3a we also invented
tasks which are either easier using a
predicative analysis or a functional one, so
that in the end we established a new type of
intelligence test. Fig. 3b shows an example
in which a predicative analysis is useful to
construct a working mental model. The main
idea is to invent a structure by arranging the
properties. One could, for example, proceed
as follows: three types of figures exist
(closed figures, figures which are open at the
Fig. 3b: QaDiPF-Example top and figures which are open at the
(Schwank, 1998)
bottom) which each have straight walls, bent
left walls and bent right walls. The figure
with an open bottom and straight walls is
missing (composition of predicates).

Fig. 3c shows an example in which
functional analysis is useful. The main idea
is that the figures in the middle row and the
middle column symbolise operators. One
could, for example, proceed as follows: in
Fig. 3c: QaDiPF-Example the first row the first figure is given thick
(Schwank, 1998) lines by means of the operator. In the first
European Research in Mathematics Education I.II: Group 5 91

column the first figure is pushed by the operator and transformed into a parallelogram.
In the second row the first figure has to be turned by means of the operator. In the last
line as a consequence the first figure has to be turned and it has to be given thick lines
(concatenation of operators).

We have designed the tasks in QuaDiPF in the form that the subjects have to
explicitly construct the missing figure instead of selecting it from a given set of
possibilities, for the following reasons: we are interested in the nature of thinking
processes and the omission of possible solutions makes the tasks more difficult.
Furthermore, we are interested in the individuality of problem-solving: a given set of
possible solutions could influence the way in which the tasks are analysed. As a
consequence our methodology is rather a qualitative one than a quantitative one.

In the literature it is discussed that solving this kind of tasks requires especially
inductive thinking (e. g. Klauer 1996). Our findings show that not only one kind of
inductive thinking exists. In a predicative model induction means abstraction. The
result is a predicate which is fulfilled by the given examples. In a functional model
induction means generalisation. The result is a function which produces the given
examples (cf. Cohors-Fresenborg & Schwank, 1996).

4. EEG-Study

Together with Jan Born and his
group, Medical University of
Luebeck, we run a study
“Dimensional complexity and
power spectral measures of the
EEG during functional versus
predicative problem solving”
(Mölle et al., in print). The EEG
was recorded in 22 young men
(Fig. 4; students at the Medical
University of Luebeck) while
Fig. 4: Phase in the EEG-Study solving QuaDiPf tasks.
European Research in Mathematics Education I.II: Group 5 92

Fig. 5: Results of the EEG-Study (cf. Mölle et al., in print)
European Research in Mathematics Education I.II: Group 5 93

Because of known gender differences in brain activities, it was important to work with
subjects of equal sex. We decided to start with male subjects, hoping that among them
will be enough typical thinkers in both modes. It turned out, that we were lucky in the
predicative case, but not the like in the functional case. Familiar with the behaviour of
the students in our department of mathematics and computer science in Osnabrueck, the
Osnabrueck staff was quite a bit astonished, that the young medical men didn’t show up
so much an ability to good functional reflections. Further experiments have to follow.
In this first experiment the subjects performed on three different blocks each including
4 QuaDiPF tasks. After having completed a pattern mentally by his own the subject had
to draw his solution and to explain why it fits the pattern well (Situation in Fig. 4). In the
first block the subjects were asked to solve the tasks spontaneously. In the second and
third block the subjects were primed to do it in a functional or a predicative way
respectively. The priming took place during three tasks, for that purpose a typical
functional or predicative argumentation for the solution was presented. The subject was
asked to solve the fourth task just in the way it had been shown to him. The EEG during
thinking on the fourth task of each condition was taken for analysis. The EEG
complexity during predicative thinking decreased in comparison to functional thinking
and mental relaxation, with this reduction being most pronounced over the right and
parietal cortex. A reduction in dimensional complexity during functional thinking as
compared to mental relaxation which was concentrated over the left central cortex,
although significant, was less clear (Fig. 5).

5. Eye Movement - Reflections

Especially tasks like 3c put some problems in adequate handling. Carpenter et al (1990)
tried to analyse the manner how subjects solve the Raven Matrices using eye-tracking
methods. They rely on a (predicative) classification of the Raven tasks, which worked
except for one task (APM No. 18 / isomorphic to fig. 3c): “Problem was not classifiable
by our taxonomy” (Carpenter et al. 1990, p. 431). This very task No. 18 is in its style
unique in the APM-Test. By means of an added functional classification on one hand
such “mysterious” tasks could be approached systematically as well and on the other
hand concerning tasks with a “double” nature like fig. 3a there could be offered,
European Research in Mathematics Education I.II: Group 5 94

additionally to the predicative classification, a functional classification. Therewith a
broader understanding of the subjects cognitive behaviour is possible.

In the near future we will run eye-tracker experiments using QuaDiPF tasks. We are
convinced, that the results will support the theory of functional/predicative thinking.
We expect, that we will found different patterns of eye-movements which are either
specific for a predicative analysis or a functional analysis: Like in Carpenter et al.
(1990) we should find eye-movements following essential properties (Fig. 6a), but
moreover we should find eye-movements along the production process of the step by
step developing states of specific objects (Fig. 6b).

Fig. 6a: Eye movement Fig. 6b: Eye movement
during predicative analysis during functional analysis

6. Outlook

When we started to run our research work we could show that the distinction between
functional versus predicative thinking is useful to analyse the behaviour of subjects in
problem solving situations in the field of mathematics and computer programming.
Viable concepts (in the sense of von Glasersfeld 1995) there can be created in one or the
other manner. Later on we could even focus on the nature of these two thinking modes
European Research in Mathematics Education I.II: Group 5 95

using much simpler tasks of a nonverbal intelligence test, which turned out to be usable
in EEG-measurement-environments.

In our days we still see the problem that it is much easier to communicate
predicative ideas than functional ones. We agree with Vandamme, that the problem will
be to find more adequate representations for functional ideas (Vandamme calls them:
action oriented, which we don’t like so much because there do exist predicative actions,
e.g. in solving a jigsaw puzzles and we want to stress the van der Waerden aspect of
motoric thinking in terms of construction instructions). We are convinced, that the new
technologies which enable to create virtual and augmented reality (e.g. Vandamme &
Morel 1996) and which bring up new tools to easily create and manage dynamic actions
on the computer screen are actually the appropriate means to master this challenge.

7. References
Bateson, G. (1980). Mind and Nature - A Necessary Unity. Toronto: Bantam Books.
Carpenter, P., Just, M. & Shell, P. (1990). What One Intelligence Test Measures: A Theoretical
Account of the Processing in the Raven Progressive Matrices Test. Psychological Review. 97, 3,
Cohors-Fresenborg, E. (1978). Learning Problem Solving by Developing Automata Networks. Revue
de phonétique appliquée, 46/47, 93-99.
Cohors-Fresenborg, E. & Schwank, I. (1996). Kognitive Aspekte des Business Reengineering.
Gestalt Theory, 18 (4), 233-256.
Cohors-Fresenborg, E. & Schwank, I. (1997). Individual Differences in the Managerial
Representation of Business Processes. In R. Pepermans, A. Buelens, C. Vinkenburg, P. Jansen
(Eds.), Managerial Behaviour and Practices - European Research Issues, 93-106.
Leuven/Amersfoort: Acco.
Klauer, K. J. (1996). Teaching Inductive Reasoning: Some Theory and Three Experimental Studies.
Learning and Instruction. 6, 1, 37-57.
Marpaung, Y. (1986): Profile indonesischer Schüler beim Umgang mit Algorithmen und ihre
Analyse. Osnabrück: Forschungsinstitut für Mathematikdidaktik.
Mölle, M., Schwank, I., Marshall, L., Klöhn, A., Born, J. (in print). Dimensional complexity and
power spectral measures of the EEG during functional versus predicative problem solving.
Brain and Cognition.
European Research in Mathematics Education I.II: Group 5 96

Raven, J. C. (1965). Advanced Progressive Matrices. Sets I and II. London: Lewis.
Schwank, I. (1986). Cognitive Structures of Algorithmic Thinking. In Proceedings of the 10th
Conference for the Psychology of Mathematics Education, 195-200. London: University of
London, Institute of Education.
Schwank, I. (1993a). On the Analysis of Cognitive Structures in Algorithmic Thinking. Journal of
Mathematical Behavior, 12 (2), 209-231.
Schwank, I. (1993b). Cognitive Structures and Cognitive Strategies in Algorithmic Thinking. In E.
Lemut, B. du Boulay, G. Dettori (Eds.), Cognitive Models and Intelligent Environments for
Learning Programming. NATO ASI Series F, Vol. 111 (pp. 249-259). Berlin: Springer.
Schwank, I. (1994). Zur Analyse kognitiver Mechanismen mathematischer Begriffsbildung unter
geschlechtsspezifischem Aspekt. Zentralblatt für Didaktik der Mathematik: Themenheft
“Frauen und Mathematik”. 26 (2), 31-40.
Schwank, I. (1995). The Role of Microworlds for Constructing Mathematical Concepts. In M. Behara,
R. Fritsch & R. G. Lintz (Eds.), Symposia Gaussiana, Conf. A., 102-120. Berlin: Walter de
Schwank (1996). Zur Konzeption prädikativer versus funktionaler kognitiver Strukturen und ihrer
Anwendung. ZDM-Analysenheft “Deutsche psychologische Forschung in der
Mathematikdidaktik”. Zentralblatt für Didaktik der Mathematik, 6, 168-183.
Schwank, I. (1998). QuaDiPF: Qualitative Diagnostical Instrument for Predicative versus
Functional Thinking. Test Set, Ver. A. Osnabrück: Forschungsinstitut für Mathematikdidaktik.
Vandamme, M. & Morel, E. (1996). Virtual and Augmented Reality: Technologies and the Manager.
Gent: The Phoenix Series.
Van der Waerden, B.L. (1954). Denken ohne Sprache. In G. Révész (Ed.), Thinking and Speaking,
165-174. Amsterdam: North-Holland.
Von Glasersfeld, E. (1995). Radical Constructivism: A Way of Knowing and Learning. London: The
Falmer Press.
Xu, B.Y. (1994). Untersuchung zu prädikativen und funktionalen kognitiven Strukturen chinesischer
Kinder bei der Auseinandersetzung mit Grundbegriffen der Programmierung. Osnabrück:
Forschungsinstitut für Mathematikdidaktik.
European Research in Mathematics Education I.II: Group 5 97

Rosetta Zan1, Paola Poli2
Dipartimento di Matematica, Pisa, Italy
Istituto di Neuropsichiatria e Psicopedagogia dell’Età Evolutiva dell’Università di Pisa
IRCCS Stella Maris , Pisa, Italy

Abstract: The aim of this study is to define the relationship between the ability in
solving mathematical problems and the beliefs about mathematical problem solving.
We compare the beliefs of children with or without difficulties in solving mathematical
problems, by using a questionnaire. The results show that good solvers and poor
solvers have significantly different concepts of a mathematical problem: some beliefs
appear to be winning in that they are able to activate the correct utilization of
Keywords: beliefs, problem-solving, school problems.

1. Introduction

Several studies in the field of the teaching of mathematics are concerned with the
emotional-motivational component of learning (Schoenfeld 1983; Lester 1987; Mc
Leod 1992). It has been shown that the failure in solving problems is not only due to the
lack of knowledge, but also to the incorrect use of knowledge which is often inhibited
by both general and specific beliefs about mathematics.

Beliefs consist of the subjective knowledge that an individual develops in attempt to
interpret the surrounding environment (Lester 1987): they influence how the subject
learns since they represent the context in which the subject selects and uses cognitive
abilities (Schoenfeld 1983; Masi, Poli, & Calcagno 1994).

Some common beliefs of general type associated with Mathematics are (Schoenfeld
European Research in Mathematics Education I.II: Group 5 98

• Mathematics problems are always solved in less than 10 minutes, if they are
solved at all.
• Only geniuses are capable of discovering or creating mathematics.

Specific beliefs about mathematics are:
• Since 31>5, then 0.31>0.5.
• The number -a is always a negative number.

Besides the beliefs associated with Mathematics, the beliefs about self are important
as well.

Borkowsky and Muthukrishna (1992) have suggested a model that relates
behavioral patterns of children facing school tasks, metacognition abilities, and self. In
their model, motivation, system of the self, development of correct learning strategies,
and self-regulation processes are closely linked. They defined a child with the
following features as a good information processor: high self-esteem, internal locus of
control (Weiner 1974), causal attribution of success and failure to effort, incremental
theory of intelligence (Dweck & Leggett 1988), and feelings of self-efficacy.

Our interest was focused on the emotional and motivational components of this
model, in particular the implicit and explicit beliefs about self - self-esteem, role of
effort, attributional theories - and learning (both the object of learning and school

More specifically, the aim of our study was to define the relationship between
solving mathematical problems, and the beliefs about self and mathematics.

To realize this we have elaborated a questionnaire (the Beliefs about Mathematical
Problems Questionnaire) with open and multiple choice questions: in the latter case the
proposed options were constructed on the basis of a previous research (Zan 1991 and
1992), that aimed at identifying, through open questions, the conceptual model of real
life problems and of school problems possessed by primary school pupils. In the course
of this research one of the following three questions was proposed to 750 primary
school pupils: “What is a problem for you?”, “Give an example of a problem.”, and
“What comes to your mind when you hear the word problem?”
European Research in Mathematics Education I.II: Group 5 99

Then we compared the answers to BMPQ given by children with or without
difficulties in solving mathematical problems. The findings presented in this paper will
concern the beliefs about mathematics, in particular the concept of a mathematical

2. Method

101 children aged from 8.5 to 10.9 participated in this study. The children were in three,
four and five grades in the primary schools in Pisa and Livorno, Italy.

The Mathematical Problem Solving Task, characterized by four standard
mathematical problems, was presented collectively in order to select two groups of
children. Children were classified in two groups: a group of good solvers composed of
30 children who solved all the mathematical problems, and a group of poor solvers of
21 children who solved one or none of the mathematical problems. Therefore, the final
sample was of 51 children.

We administered the Beliefs about Mathematical Problems Questionnaire to
investigate the concept of a mathematical problem. Questions were read aloud by the
examiner and there was no time limit to answer.

3. Results

Owing to lack of space, we discuss in detail only the answers to the most significant
questions. In particular, we omit the answers to the open questions and the
justifications for the multiple choice questions, as they would require a more complex

For simplicity, we have inserted after each question the associated table with the
relative data and comments. In order to decide whether the difference between the two
groups (good solvers / poor solvers) is significant, we have collected together the
answers to some of the questions, naturally on the basis of theoretical criteria.
European Research in Mathematics Education I.II: Group 5 100

3. In your opinion, why are mathematical problems called problems?
(Choose just an answer)
[A] This is a usual word to call them: they might also be called “exercises”.
[B] Because for the mind there is a difficult situation to solve.
[C] Because for a child who is unable to solve it, it becomes a problem.
[D] Because they describe someone’s problem and we are asked to solve it.

Question 3 aimed at deciding whether the children reduce the concept of
mathematical problem to the more general concept of real problem: this is possible in
several ways (answers B, C, D). In case C, unlike B and D, the problematic situation is
external to the task and centered on the subject. From a different point of view, the
children who choose B recognize as problems a broader variety of situations in
comparison with those who choose D. These remarks are summarized in the following

Question 3 - Real pr.
A No Involvement
C Yes ego-centered Generality
D Yes task-centered low
B Yes task-centered high

In our opinion, the answers A and C correspond to less effective approaches.
Because of their small number, they are collected in one group.

Most good solvers choose answer B, whereas most poor solvers choose answer A or
C. [chi2=6.153; p=0.047]

Question 3 Good solvers Poor solvers
A or C 6 11
B 17 6
D 7 4
European Research in Mathematics Education I.II: Group 5 101

4. What is a mathematical problem?
[A] It is a text with some numbers and a question.
[B] It is a situation that you can solve by using mathematics.
[C] It is an exercise where one has to decide which operations should be done and
then do them.
[D] It is an exercise presented during a mathematics lesson at school.

Question 4 Good solvers Poor solvers
[A] 4 4
[B] 23 5
[C] 2 6
[D] 1 6

Most good solvers choose answer B, whereas most poor solvers choose answer A or
C or D.
[chi2=13.939; p<0.001]

5. Does there exist a mathematical problem without numbers?

Question 5 Good solvers Poor solvers
Yes 21 8
No 9 13

[chi2= 5.126; p=0.024]

9. Alessandro says: “A problem with many questions is more difficult than a
problem with one question.” Do you agree with him?

Question 9 Good solvers Poor solvers
Yes 6 11
No 24 10

[chi2= 5.828; p=0,018]
European Research in Mathematics Education I.II: Group 5 102

10. Alice says: “A problem with a short text is easier than one with a long text.” Do
you agree with her?

Question 10 Good solvers Poor solvers
Yes 3 13
No 27 8

[chi2=13.8 with Yates correction; p< 0.001]

28. In a problem is it worse to make a calculation error or to choose the wrong
[A] Calculation error.
[B] Choose the wrong operations.
[C] It’s the same, there is no difference.

Question 28 Good solvers Poor solvers
[A] 8 15
[B] 20 5
[C] 2 1

[Considering only answers A and B, chi2=8.303; p=0.004]

34. How do you feel when the teacher says: “Now let’s do a problem.”?
[A] You are excited but happy.
[B] No particular feeling.
[C] You are nervous because you don’t know if you will be able to solve it.
[D] You are quite scared.

Question 34 Good solvers Poor solvers
[A] 14 2
[B] 11 1
[C] 1 5
[D] 4 13
European Research in Mathematics Education I.II: Group 5 103

We have collected together the answers corresponding to positive or neutral
emotions (A and B), and those corresponding to negative emotions © and D).
[chi2=21.649 with Yates correction; p<0.001]

4. Conclusions

We distinguished four categories of children based on the analysis of the responses
given to question 4 of the questionnaire:

1. Formalists - they recognized the mathematical problem on the basis only of the
formal features of the text [i.e., it is a text with some numbers and a question];
2. Structuralists - the mathematical problem was identified by using mathematical
tools [i.e., it is a situation that you can solve by using mathematics];
3. Operatives - the presence of arithmetic operations defined the mathematical
problem [i.e., one has to plan the arithmetic operation and do it];
4. Pragmatist - for these, the mathematical problem was characterized by contextual
elements [i.e., it is presented during a mathematics lesson at school].

The good solvers significantly belonged to the category of structuralists.

• The motivation of good solvers was focused on the task; on the contrary, the
motivation of the poor solvers was self-centered;
• The poor solvers were more sensitive to syntactic cues - text length, number of
questions, magnitude of numbers - than the good solvers in judging the difficulty
of a mathematical problem;
• The poor solvers considered an error in calculation worse than one in planning or
selecting the correct arithmetic operations;
• The poor solvers were aware of being anxious when facing a mathematical

In conclusion the good solvers and the poor solvers have a significantly different
concept of a mathematical problem.
European Research in Mathematics Education I.II: Group 5 104

The difference between the answers of the two groups suggests the following
definition. We call “winning” the beliefs of the good solvers, because these beliefs
appear to be able to evoke the correct utilization of knowledge.

Metacognitive research has shown that, through a direct intervention, the beliefs
about self can be modified. In this way, the cognitive and metacognitive resources can
be activated.

Hence, in view of this research, our results suggest a teaching approach to
difficulties in problem solving that makes children’s beliefs explicit and endeavours to
modify them.

5. References
Borkowsky, J.G., & Muthukrishna, N. (1992): Moving motivation into the classroom:"Working
models" and effective strategy teaching. In M. Presley, K.R. Harris, and J.T. Guthrie (Eds.)
Promoting Academic Competence and Literacy in Schools. Academic Press, San Diego.
Dweck, C.S., & Leggett, E.L. (1988): A social-cognitive approach to motivation and personality.
Psychological Review, 95 (2), pp. 256-273.
Lester, F.K.Jr. (1987): Why is problem solving such a problem? In Proceedings PME XI, Montreal.
Masi, G., Poli, P., & Calcagno, M. (1994): Metacognizione e apprendimento della matematica. In
C.Caredda, B.Piochi, and P.Sandri (Eds.) Handicap e svantaggio. Pitagora, Bologna.
Mc Leod, D. (1992): Research on affect in Mathematics Education: a reconceptualization. In Douglas
A. Grows (Ed.), Handbook of Research on Mathematics teaching and learning. Macmillan
Publishing Company, New York.
Schoenfeld, A.H. (1983): Theoretical and Pragmatic Issues in the Design of Mathematical Problem
Solving. Paper presented at The Annual Meeting of the American Educational Research
Association, Montreal.
Schoenfeld, A.H. (1985). Mathematical Problem Solving. Academic Press, Orlando, FL.
Weiner, B. (1974): Achievement motivation and attribution theory. Morristown, New York: General
Learning Press, 1974.
Zan, R. (1992): I modelli concettuali di problema nei bambini della scuola elementare.
L’insegnamento della matematica e delle scienze integrate, vol. 14, n.7 and n.9, 1991, pp.
659-677 and 807-840, vol.15, n.1, pp.39-53.

European Research in Mathematics Education I.II: Group 6 106

coordinated by Paolo Boero
Department of Mathematics, Genoa University

1. General Information

Pre-conference preparation was coordinated by C. Bergsten (Sweden), P. Boero (Italy)
and J. Gascon (Spain). Ten contribution proposals were received, eight of which were
accepted, distributed to all participants before the conference and discussed during the
conference. Coordination of Working Group activities was ensured by C. Bergsten and
P. Boero. From ten to twelve participants took part in the WG sessions; nine of them
were present at all sessions. Countries represented within the WG were France,
Germany, Hungary, Italy, Spain, Russia and Sweden.

2. Methodology

Given the content of the papers, it was decided to distribute the five time slots at our
disposal as follows:

• three hours for the topic “The nature of algebra related to the teaching of algebra”
(discussion mainly concerned the papers by M. Bosch & al and S. Rososhek);
• an hour and a half for “Arithmetic and algebra in schools” (paper by N. Malara);
• three hours for “Tools for research in the teaching and learning of algebra”
(papers by C. Bergsten and by E. Szeredi & J. Torok);
• an hour and a half for “Equations” (paper by M. Reggiani);
• three hours for “Construction and interpretation of algebraic expressions related
to other domains” (papers by L. Bazzini and by V. Hoyos & al).
European Research in Mathematics Education I.II: Group 6 107

The papers discussed in each time slot were intended as starting points for the
discussion. In reality, many links were made between the first and the third topics, and
between them and the other topics. The discussion of papers usually started with
previously prepared reactions from some members of the group. The papers were not
presented during the WG sessions; however, during the discussion of a paper, the
author was sometimes asked to provide “local” information (examples, explanations,
etc.) or to contextualise the content of that paper (especially in the case of pieces of
research belonging to wide-ranging projects).

3. Summary of Discussions

The following summary is organised according to the topics listed above; it outlines a
few points discussed during the sessions of the working group, in order to give the
“flavour” of the debate. It is interesting to note that some contributions on one topic
came from discussions related to another topic.

The summary has been checked and improved by the members of the WG.

3.1 The Nature of Algebra Related to the Teaching of Algebra

During our discussions we considered different “definitions” of algebra:
• “absolute definitions”, i. e. not specifically related to teaching situations; for
instance “study of algebraic structures”, “modelling of mathematical activities”,
“specific activities related to the use of algebraic language”, etc.
• definitions concerning “school algebra”: for instance, “generalised arithmetic”,
“mastery of algebraic language”, etc.

Each work in mathematics education concerning algebra is based on a specific
epistemological model of algebra, even if sometimes the definition used is not
explicitly presented or is taken for granted. The particular definition of algebra that is
assumed has different implications for the teaching and learning of algebra in schools
(aims, content, methods). For instance, the following aims of the introduction of
algebra in the curriculum are related to the Weyl-Shafarevich’s conception of algebra:
European Research in Mathematics Education I.II: Group 6 108

to overcome syntactical difficulties in the mastery of algebraic expressions; to develop
students’ imagination in mathematics; and to enhance metacognitive processes.

It was noted that it is useful for research in mathematics education to consider
different models of what algebra (and learning algebra) is, and to “play the game”
within the chosen models. As an example, algebra was considered a process of
modelling a whole mathematical work (which includes types of problems, different
kinds of techniques to solve these problems, and theoretical tools to “talk about”, justify
and expand the resolution of these problems). According to this perspective we can say
that the introduction of algebraic tools needs a previous construction of a whole
mathematical activity (be it numerical or geometrical). With this “anthropological
definition of algebra”, we can also show some mathematical and didactic restrictions
that limitate the introduction of algebraic tools in schools and can explain why
secondary school curricula are only partially algebraized. Nevertheless it was remarked
that this way of considering algebra is weak on the important symbol transformation
aspects of algebra. Another example viewed algebra as a matrix of the dimensions
“content” vs “activity”, i.e. equations, generalisations, and relations vs translating,
transforming, and interpreting. This model is close to school algebra practice, but is
weak from the meta level perspective, since it is more describing than analysing the
whole setting.

3.2 Arithmetic and Algebra

If we consider “school algebra” not as “generalised arithmetic” but as a systemic
activity related to the use of the algebraic language, some activities in the field of
elementary arithmetic can be performed in an algebraic perspective (“teaching
arithmetic in an algebraic mode”). As an example, we may consider (in the first grade)
the activity of representing the number 7 in different ways: 7=4+3=6+1. But this kind of
activities has raised an interesting issue: the teacher’s conceptions about algebra as a
possible source of obstacles in the direction of an arithmetic curriculum partly oriented
in an algebraic perspective. As concerns this perspective, teacher’s difficulties are not
local (i.e. related to lack of some specific notions), they are global. So the aim of early
introduction of basic algebra skills in primary school must take account of teachers’
European Research in Mathematics Education I.II: Group 6 109

pre-service training and conceptions and their possible evolution through suitable
in-service training based on reflection about classroom experiences related to the
nature of the discipline.

3.3 Tools of Research in Teaching and Learning of Algebra

Different theoretical frameworks and tools belonging to various disciplines related to
mathematics education were extensively presented during the discussion by people
which currently use them in their research work: two specific theoretical frameworks in
the area of mathematics education:

• Brousseau’s “Theory of didactical situations”;
• Chevallard’s “Anthropological theory”;

and other frameworks and tools from different disciplines:

• Johnson & Lakoff’s “image schematas”;
• other tools belonging to cognitive psychology;
• Weyl-Shafarevich’s epistemological position about algebra;
• teaching and learning implications of Frege’s distinction between “sense” and
“denotation” of a sign;
• epistemological obstacle (as an interpretative tool for students’ and teachers’

Two important and delicate questions were repeatedly tackled: What research
problems in the teaching and learning of algebra are related to what theoretical
frameworks? And is there a difference between specific (to mathematics education)
and non specific (taken from other disciplines) frameworks?

Discussion concerned the “generative” character of theoretical frameworks in the
definition of research problems and of teaching and learning phenomena.
European Research in Mathematics Education I.II: Group 6 110

Another (meta-)research question concerned the possible need for complementarity
of tools belonging to different disciplines and theoretical frameworks in order to
investigate specific aspects of the teaching and learning of algebra. For instance, the
misuse of the symbol S by a student during the solution of a “generalization and proof”
problem was interpreted from the anthropological point of view (considering the use of
S as a part of a mathematical technique that might have been unavailable to the student)
and the cognitive point of view (considering the difficulties inherent in the mental
processes involved in the use of S ).

The difficult problem of validation of research tools was also raised.

Finally, the question of the validity of “comparative studies” concerning the
teaching and learning of algebra was discussed. If the many relevant variables are taken
into account, it seems very unlikely at this moment that valid comparative studies can
be done. Nevertheless, comparisons might be very useful: they might suggest
interesting research issues, research hypotheses, etc.

3.4 Equations

Two research questions were focussed on:
A) aims (and concrete possibility) of early reflective activities about equations (e.g.
discussing the equivalence of 0.x=5 and 0.x=10); must reflective thinking come
necessarily after action? Must the concept of equation be extensively practised (by
solving standard equations), before reflection be started?
B) what theoretical frameworks (in the field of didactics of mathematics and in the field
of cognitive psychology) are needed to tackle the preceding issue? For instance, a
curriculum based on the relevance of Frege’s distinction between “sense” and
“denotation” of a sign tends to bring forward the activities indicated in A. And
analysis of the difficulties (met by students in tackling those activities) in terms of
the “didactical contract” points out the break between the arithmetic and algebraic
approaches to equations. Indeed the algebraic and the arithmetic perspectives imply
an important change in the meaning of the same activities (for instance, solving an
European Research in Mathematics Education I.II: Group 6 111

3.5 Construction and Interpretation of Algebraic Expressions Related to
Other Domains

Some research questions emerging from the discussions were:
• the nature of the difficulties in the transition from verbal expressions (related to
arithmetic, geometry, etc.) to algebraic expressions. Different interpretative
hypotheses were suggested, some related to semiotics (different organization of
verbal and algebraic languages), others to cognitive psychology (based on the
different image schematas involved), and others to didactics of mathematics
(familiarity with different kinds of techniques; familiarity with this kind of
“translation” task);
• potentialities (and possible limitations and ways to overcome them) inherent in
the dynamic environment of Cabri software in the transition from geometric
construction of curves to their algebraic expressions as equations. The possibility
of easy “trials” could induce students to avoid the effort of elaborating suitable
algebraic expressions through a deep analysis of the curve. But this danger could
be avoided through a suitable choice of tasks related to the specific
characteristics of the software.

4. General Comments

The “circumscribed” topic and the sound scientific basis of the accepted papers
produced discussions that were very productive in terms of communication and mutual
understanding. These two aspects prevailed over the collaboration perspective; perhaps
no other outcome was possible given the divergence in the research traditions and
frameworks represented. However, some hints for collaboration did emerge: for
instance, testing different research tools on the same “corpus” of protocols; and
comparing different approaches to algebra (all based on preliminary activities in the
domain of arithmetic).
European Research in Mathematics Education I.II: Group 6 112

Luciana Bazzini
Dipartimento di Matematica, Università di Pavia, Via Abbiategrasso 215, I-27100 Pavia, Italy

Abstract: Recent research studies have pointed out the crucial role of constructing and
interpreting letters in algebra. Many difficulties emerge because of the incapability to
relate the algebraic code to the semantics of the natural language.
A teaching experiment, carried out with 16 year old students, attending the second year
of the Gymnasium (a humanistically oriented High School) is described here. This
experiment was aimed at analysing the cognitive behaviour of the students when facing
learning situations dealing with a productive use of symbols and their understanding.
Keywords: algebra, natural language, symbols.

1. The problem’s Position

The general consensus considering symbols as a driving force of algebraic thinking has
fostered a deeper investigation on the dialectical relationships between signs and
algebraic ideas as well as a growing interest for teaching implications

There is evidence that the use of symbolic expressions can be a relevant cause of
difficulties in students (see for example a study by Radford and Grenier, 1996).
According to Laborde (1982) the development of a specific symbolic language can
impoverish the meaning of the language previously used. If we look at the historical
evolution of algebra, we see that rhetoric and syncopated algebra (i.e. algebra totally
expressed by words and algebra expressed by a mixture of words and symbols) have
been quite easy to use and understand. On the contrary, in a symbolic system the
meaning of words and operations can stay behind the scene, since the symbolic
language has the power of taking away most distinctions that are preserved by the
natural language. Because of this specificity, the symbolic language expands its
applicability but induces a sort of semantic weakness: it seems that such language is
European Research in Mathematics Education I.II: Group 6 113

suitable for many contexts, without belonging to anyone. Hence the origin of the gap
between symbols and meanings, which is confirmed by the rigid (stenographic) use of
the algebraic code by many students. It often happens that some students are able to
control the underlying meaning only when they make use of rhetoric or syncopated

The “rhetoric” method seems to be used spontaneously (Harper, 1987) and does not
depend on the instruction level.

Research studies dealing with the process of constructing and interpreting letters
(variables or parameters) point out its crucial role in algebra.

In this perspective the choice of names to designate objects is strictly related to the
control of the variables which are involved to characterize the properties to be
emphasized (Arzarello, Bazzini, Chiappini, 1994c). Difficulties emerge because of the
impossibility to relate the algebraic code to the semantics of the natural language. The
student able to express the relationships among the elements of a given problem
correctly, by means of the natural language, can be unable to express the same
relationships through the algebraic code.

The process of construction and interpretation is sometimes impoverished or
blocked when the subject considers the terms in a rigid way, and does not grasp the
underlying interrelation between sense and denotation of a given name (Arzarello,
Bazzini, Chiappini,1994a,1995). In short, there is evidence that the student is often not
able to take the whole potential of the algebraic code, i.e. the power of incorporating
different properties within the name. The name is seen as a rigid designator, source of
obstacles for algebraic thinking. Consequently, growing difficulties appear in front of
algebraic transformations, and their additional requirement of foreseeing and applying,
guessing and testing the effectiveness, in a continuous tension (Boero,1994).

All these issues foster a careful analysis of the questions related to the learning of
algebra as a language; such questions are rooted, at an early school level, in the dialectic
relation between semantics and syntax, procedures and structures, natural and symbolic
language. The passage from natural language to symbolic language is a key point in the
development of algebraic thinking and asks for special attention in teaching.
European Research in Mathematics Education I.II: Group 6 114

Furthermore, we have to take into account the outlined questions framed in the wider
perspective of the use and abuse of mathematical symbols in school practice .
Sometimes students’ spontaneous symbols are not encouraged in school to develop
towards generally accepted symbols. Hence, symbols are often the cause of many
learning difficulties, because of their loss of meaning (see for example a study by
Furinghetti and Paola, 1994).

2. The Teaching Experiment

A teaching experiment, carried out with 16 year old students, attending the second year
of the Gymnasium (a humanistically oriented High School) is described here. The
experiment was carried out in the period March-June according to the usual school year
schedule: no additional lessons have been given to students. The experiment included a
training phase (approximately three months), a break (two weeks) and a final test
consisting of a questionnaire and clinical interviews.

The class was made up of 20 students at upper intermediate level. The mathematics
teacher and an interviewer have been present in the classroom during the whole
experiment, which was centred on the topic “Straight lines and linear systems”.

This experiment aimed at analysing the cognitive behaviour of the students when
facing learning situations dealing with a productive use of symbols and their

The main objective is the study of the difficulties which persist in mastering
symbols (namely the construction and interpretation of symbolic expressions) for
students who have received a properly oriented training.

Since the previous school year, the teacher of mathematics had systematically been
practising a meaningful oriented teaching, with special attention to the role of symbols.
In each lesson a careful analysis of the symbolic expressions encountered have been
done: the students have been requested to distinguish between unknowns, variables and
parameters. They have also been required to reflect on the changes induced by the
variation of a given letter.
European Research in Mathematics Education I.II: Group 6 115

At the end of the experiment, a questionnaire was administered to the students and,
finally, each student was interviewed.

For our purposes, we will focus on two items of the questionnaire and two questions
of the interview. Let’s notice that the questions proposed (in the questionnaire as well
as in the interview) are not routine questions, because they should test a genuine
mastery of the symbolic code.

Here following the text of the two problems in the questionnaire:

Given the straight lines y=3x+q and y=mx+5
a) give a condition for being parallel
b) give a condition for being perpendicular
c) give a condition for having the point (1,3) in common
d) give a condition for having the point (0,0) in common.

A linear system of two equations and two unknowns, admits the couple (1,1) as a solution.
Discuss the following statements:
To get a solution which is double of the given one, (i.e. to get the solution (2,2)), one
a) multiply both sides of the two equations
b) multiply by two both sides of one equation
c) construct a different system.

2.1 Problem I: Analysis of the Students’ Answers

Points a) and b) were correctly solved by 16 students (17 being the total number of
students): just one did not give any solution.
The correct solutions at point c) were 10 (3 omitted and 4 wrong).
At point d) the correct solutions were13 (one omitted and two wrong).
There is evidence that the great majority performed well in this task.

As far as point c) is concerned, two students assigned arbitrary values to m and q,
one worded a correct procedure but did not provide the numerical solution, another
carried out a sequence of algebraic transformations unsuccessfully.
European Research in Mathematics Education I.II: Group 6 116

At point d) a student replied:”No straight line passes through the origin”: it seems
that the presence of the letter q “automatically” implies that the known term is different
from zero. This misconception is not unusual: it appeared in similar conditions along
the year. Other two students hold the belief that the straight line could pass, in some
way, through the origin. More precisely, one student said:”If m=5, then the equation
y=mx+5 becomes y=x, i.e. a straight line through the origin” Another student
claimed:”The equation y=3x+q passes through (0,0) when q=0; when the other
equation too (i.e. y=mx+5) will have q=0, the two lines will meet in the origin”.

These kinds of responses clearly point out the presence of rigid designators (q
different from zero and the possibility of manipulating formulas to be adapted as

2.2 Problem II: Analysis of the Students’ Answers

This problem was correctly solved by all the students. This confirms the good mastery
of the algebraic meaning underlying a linear system and the operations which do not
change the system’s solution. Nine students out of 17 needed numerical trials. Five
students used the geometrical referent: two replied that one needs to apply the
translation (1,1) to both lines; three justified the incorrectness of the first two answers
referring to the linear combination of the two equations: by multiplying the two
equations by different factors, one gets lines still passing through the centre of the

Let’s consider now two of the questions administered during the interviews:

I) The cube problem
Try to express in an algebraic language that the ratio between the volume of a cube,
whose edge is unknown, and the area of the square, whose edge is the double of the
cube’s edge, is two times the area of one face of the cube, which is augmented by three.

II) Which sense the following expressions have for you:
x2=px+q y=mv+h
European Research in Mathematics Education I.II: Group 6 117

2.3 Question I): Analysis of the Students’ Answers

Just five students out of 19 solved the question easily. More precisely, just one
performed well, without any help: the others had some slips (i.e. 2l2 instead of 4l2 or
l3/4l2=2l2+3 instead of l3 /4l2=2(l2+3). The remaining 14 students needed to read the
text several times: probably they would not have been able to approach the problem
without any help from the interviewer. Difficulties have been of a different nature:
some due to deficiency in computation (for instance “ss=2s”) or in the symbolic
translation of simple sentences expressed in the natural language. For example a
student asked if “to augment by 3” would mean “to add 3” or “to multiply by 3”.

A student wrote “Vc/Aq =3(2Ar)”; another one approached the question by recurring
to a numerical support, i.e. by attributing an arbitrary value (2) to the length of the
cube’s edge (and being consequently 4 the length of the edge of the square). Here the
report of such trial: “I have a cube of four little squares, the cube has six faces, hence
4x6, then I have another one with doubled edge ...”

“I have to compute the ratio of these two, i.e. 4x6.........”
The interviewer underlines that they are speaking of a volume, and then she writes 4
and 2.

The interviewer asks “Why 2 and 4?” and she replies that the area of the square is l2,
the area of the cube’s face is ½ l2. She continues without any visible orientation:
European Research in Mathematics Education I.II: Group 6 118

½ l/l=2l+3 (½ l )* (½ l )=2l + 3 ¼ l = 2l + 3. and does not reach any plausible

In this episode we observe that numbers are intended to support the formula to be
constructed, but the trial fails.

Many students made the mistake of not using the parenthesis in the second side of
the equation: 14 students wrote “... =2l2+3” instead of “...=2(l2+3)”. Nine students
expressed “the area of the square having the edge doubled with respect to the edge of
the cube” as “2l2” instead of 4l2 and they changed their writing only after the
intervention of the interviewer.

Eight students did not express “the area of a square having the edge doubled with
respect to the edge of the cube” or “the area of the face of the cube” as function of the
cube’s edge, but used several unlinked symbols. Here are some examples:

S3 / 2S2 = 2a2 +3

spig3 : l2 =2 (spig2 +3)

V/(2l)2 = 2(l2)+3

Vc / Aq = 2Af +3

A c = 2lr 2lr

These students created a new symbol (sometimes even figurative, i.e. A c ; 2lr ) for
each new element mentioned in the text, showing thus a very poor mastery of the
synthetic function of the algebraic language. Finally a student wrote a totally out of
focus expression: p h / 4p = 2p +3

During the interviews, great difficulties emerged as far as the reading of the text was
concerned: some students were able to memorize only a part of the text. Such difficulty
is rooted in the incapacity to consider all things together: when a student forgets to
European Research in Mathematics Education I.II: Group 6 119

consider a part of the problem, there is evidence of the difficulty in naming, which
makes incorporating and clarifying the sense of a problem into the system of algebraic
representation possible (see Arzarello, Bazzini, Chiappini, 1994b).

2.4 Question II): Analysis of the Students’ Answers

The expression x2=px+q was considered as the equation of a straight line by 8 students
out of 19. Other eight students interpreted the given expression as a second grade
equation. One student replied. “It does not remind me of anything”: after the
intervention of the interviewer, he recognized that it could be an equation in p or q and,
hence, a straight line. Just one girl gave a complete answer: “It could be a literal
equation of second grade in x, or an equation in the unknowns p and q, where x and x2
are the known terms, or an equation in q (the only unknown), x2 and px being the known
terms”. Another student interpreted the expression in terms of objects: “A squared
thing, which equals itself multiplied by another thing, plus another additional thing”.

Eight students interpreted the expression as the equation of a straight line: one of
them claimed “It is a squared straight line”. Four students specified that they were not
able to justify the presence of x2, but this was not enough to contrast their wrong belief.

Eight students interpreted x2=px+q as a second grade equation: two of them noticed
that if x2 would have been substituted by y, they would have recognized a straight line.

This confirms the relevance, in the students’ minds, of the pattern of straight lines.

Finally, an unusual behaviour of two students: they considered x2=px+q as the
equation of a straight line in the unknowns p and q.

The expression y=mv+h was considered by 9 students as the equation of a straight
line, by eight students as a first grade equation.

Two students made reference to a formula constructed in a different context (not
related to algebra or geometry), for example physics or chemistry.
European Research in Mathematics Education I.II: Group 6 120

It is worth noticing that five of the nine students who interpreted the expression as
the equation of a straight line, clearly state that they were considering the expression as
if v were substituted by x and h by q. Such students were flexible in seeing a straight line
in the given equation, but they needed to refer to the standard writing y=mx+q (rigid

As observed, eight students saw a first grade equation: most of them considered the
y as unknown. One student said it was a first grade equation in x (which did not appear):
again the sign x as a rigid designator of the unknown.

3. Final Remarks

The experiment points out some key issues in the process of constructing and
interpretating algebraic expressions.

It is clear that many problems are rooted in the students‘ incapability to construct
suitable formulas, which should incorporate the meanings of the objects involved and
their mutual relationships. This process is typical of algebra and constitutes a breaking
away from arithmetic: it is something new to students, and even in the historical
development of the discipline.

The experiment has shown that even bright students, who have received a good
teaching, (that is oriented towards a meaningful understanding of symbols) and have
performed well in the above mentioned problems of the questionnaire, encounter
relevant obstacles when passing from rhetoric (or syncopated) algebra to symbolic
algebra. In the change many students do not grasp the symbolic function of the
algebraic language. As Norman (1987) pointed out, since the algebraic language is not
directly associated to the semantics of the natural language, this latter can influence and
distort the construction of formulas.

In fact, the choice of names to designate objects is linked to the control of the
introduced variables: difficulties emerge because it is very unusual that algebraic
formulas are a simple linear stenography of what is expressed by means of the natural
European Research in Mathematics Education I.II: Group 6 121

Consequently, the change from one semiotic register to another (natural language
and symbolic language) does not occur appropriately and knowledge is not constructed
(Duval, 1995).

As noticed by many authors, much work towards solution is done when a good
process of naming has been produced by the subject (see, for example
Chalouh&Herscovics, 1988; Kieran, 1989; Arzarello, Bazzini, Chiappini, 1994b).

The experiment has given further evidence of the students’ difficulties in naming the
elements of the problem in an appropriate way and coaching their stream of thinking in
order to “condense” the most relevant aspects of the problem into a formula.

These persistent difficulties confirm our belief that learning algebra should be the
result of a long cognitive apprenticeship during which the student learns to incorporate
senses in algebraic terms and expressions, overcoming the stereotypes of rigid
designators and eventually controlling the senses of the constructed expressions.

4. References
Arzarello F., Bazzini L., Chiappini G. (1994 a): Intensional semantics as a tool to analyze algebraic
thinking, Rendiconti del Seminario Matematico, Univ. Torino.Vol. 52, N.2, (105-125).
Arzarello F., Bazzini L., Chiappini G.P. (1994b): The process of naming in algebraic problem solving,
Proceedings of PME XVIII, Lisbona, Vol.II, (40-47).
Arzarello F., Bazzini L., Chiappini G. (1995): The construction of algebraic knowledge: towards a
socio-cultural theory and practice. Proceedings of PME19, Recife, Brazil, Vol I , (119-134).
Boero P. (1994): About the role of algebraic language in Mathematics and related difficulties.
Rendiconti del Seminario Matematico, Univ. Torino.Vol. 52, N.2, (161-194).
Chalouh L., Herscovics N. (1988): Teaching algebraic expressions in a meaningful way, in Coxford
A.F., The ideas of Algebra, K-12, 1988 Yearbook, NCTM, Reston Va.
Duval R. (1995): Sémiosis et pensée humaine. Peter Lang, Bern.
Furinghetti F., Paola D. (1994):, Parameters, unknowns and variables: a little difference? Proceedings
of PME XVIII, Lisbona, Vol.II, (368-375).
Harper E. (1987): Ghosts of Diophantus, Educational Studies in Mathematics, Vol.18, (75-90).
European Research in Mathematics Education I.II: Group 6 122

Kieran C. (1989): The Early Learning of Algebra, A Structural Perspective, in Wagner S., Kieran C.
(Eds.) Research Issues in the Learning and Teaching of Algebra, NCTM, L.Erlbaum Ass.,
Hillsdale, N.J.
Laborde C. (1982): Language naturelle et écriture symbolique, These, Université J.Fourier,
Norman F.A. (1987): A psycholinguistic perspective of algebraic language, Proceedings of PME11,
Vol.I, (234-230).
Radford L., Grenier M. (1996): On dialectical relationships between signs and algebraic ideas,
Proceedings of PME XX, Valencia, Vol.4, (179-186).
European Research in Mathematics Education I.II: Group 6 123

Christer Bergsten
Department of Mathematics, Linköping University, S-58183 Linköping, Sweden

Abstract: In this theoretical paper figurative aspects of algebraic symbolism are
discussed and related to the theory of image schemata, opening up for one way of
understanding the development of symbol sense in mathematics. The ideas of
mathematical forms (referring to spatial characteristics of mathematical formulas),
and form operations, are at the core of this analysis.
Keywords: -

1. Introduction

The term ”symbol sense” has only recently come into use in mathematics education,
referring to a similar kind of familiarity with algebraic symbols as ”number sense” to
arithmetic: ”As students’ understanding of algebra deepens, they are gaining symbol
sense: an appreciation for the power of symbolic thinking, an understanding of when
and why to apply it, and a feel for mathematical structure. Symbol sense is a level of
mathematical literacy beyond number sense, which it subsumes.” (Picciotto & Wah
1993, p. 42) With a developed symbol sense as the main goal of algebra teaching
(Arcavi 1994), a theoretical framework of the conception and its development is
needed. The issue is complicated by the fact that algebra can be viewed both as a
symbol system and as a way of thinking (cf Sierpinska 1995, p. 157). The interest here
is on aspects of the symbolic language used in school algebra.

Leibniz, one of the most influential innovators of notations in the history of
mathematics, wrote (see Cajori 1929, p. 184): “In signs one observes an advantage in
discovery which is greatest when they express the exact nature of a thing briefly and, as
it were, picture it; then indeed the labour is wonderfully diminished.” This quotation
(also quoted in Bergsten 1990), referring to notations in mathematics, indicates that
when symbols in mathematical formulas are arranged in a way similar to those
structures of the real world that the formulas are used to depict, then the notation is easy
European Research in Mathematics Education I.II: Group 6 124

to understand and to work with. The idea that mathematical thinking, as well as
understanding, as related to formulas, can be facilitated by observing the figurative
characteristics of formulas, can find further support in the image schemata theory of
Mark Johnson and George Lakoff, as noted by Dörfler (1991). In this theoretical paper
some conceptions of mathematical form, introduced in Bergsten (1990), and in line
with these ideas, are further developed and related to symbol sense. It should be pointed
out that the space available for this conference paper does not allow a deeper
elaboration of the ideas presented.

2. Mathematical Forms

When looking at strings or expressions of (standard) mathematical symbols (in
elementary school mathematics) as pictures, only a very few basic patterns (spatial
arrangements) appear. Bergsten (1990) thus defined forms as typographical units
(atomic forms or symbols) or as spatial relations between typographical units
(molecular forms or patterns). The notion of mathematical forms refers to symbols
used in mathematics and patterns of such symbols. When combining atomic
mathematical forms like 1, 2 and 3 to produce mathematical expressions like 213, 1+2,
or 1+2=3, the molecular forms appear as the spatial characteristics (schematic
structures) of the symbolic expressions. The forms here can be depicted by the
schematic structures of the figures III, IOI, and IOIOI respectively. Seeing 1+2 as a unit
(chunking) in the last of the expressions, the essential form is seen to be IOI there as
well. This is the form of a link (that can be extended to a chain like in IOIOI), used in
notations of both operational (1+2) and relational (1<2) ideas. The other molecular
form that appeared here (i e III) has the pattern of bars. An algebraic expression like 3x
+ y = 5 is a molecular form structured by the following “Chinese boxes”: ( ( ( 3·x ) + ( y )
) = ( 5 ) ). It can be observed that the pattern is composed by a number of links only, in
this case three.

Also when operating on mathematical symbols some basic schematic structures
appear. Some of these are splitting (as in 3x ® x+2x), joining (x+2x ® 3x), mirroring
or exchange of position (b+a ® a+b, adding (3 ® 30 when multiplying by ten),
deleting (30 ® 3 when dividing by ten), raising (2x ® x2 when integrating), lowering
European Research in Mathematics Education I.II: Group 6 125

(x2 ® 2x when differentiating). The schematic structures describing these symbol
transformations can thus be seen as form operations and are composed into form
operators (transforming for example an equation like 7x-3=4x+6 into its solution x=3).
Also the identity operation can be identified as a form operation. The role of form
operators in algebra is essential. Pimm (1995) even argues that ”the algebra takes place
between the successive written expressions and is not the statements themselves.” (p.

Some basic forms seem to have their origin in bodily experience. These forms can
be called genetic (Bergsten 1990), by which is meant that they can be isomorphically
mapped onto pictures (or objects) of possible referents of the symbolic expressions. A
simple example is provided by the expression 2+3, the form of which (link) also
appears in the spatial arrangement of the picture oo ooo , showing the act or the idea of
adding together two and three objects, a possible referent of the symbolic expression.
The form of the algebraic counterpart a+b is inherited from the world of arithmetic, thus
bringing along its genetic character. Forms that are not genetic are called stipulated, as
most atomic forms (like 2 or Ö) or the form superposition used in exponential notation
(as in x2 or ex).

Observations like these are of course trivial but seem to have far reaching
implications for the understanding of (some parts of) mathematical symbolism:
• The three molecular mathematical forms just mentioned are not only
commonplace, but practically the only ones that are used in (standard)
elementary school mathematics to formalize the ideas of numbers and the four
basic operations of arithmetic (forms for executing some algorithms excluded).
They are also (notably as a consequence) commonplace in algebraic notation
(including abstract algebra and vector algebra).
• The observation that genetic forms seem to be based on bodily experience invites
the use of image schemata theory in discussing understanding and meaning of
(some parts of) mathematical symbolism. In particular, the individual
development of (some aspects of) symbol sense can be traced back to the
structure of abstract thinking ultimitaley based on sense impressions.
European Research in Mathematics Education I.II: Group 6 126

• A structure-operation distinction within mathematical symbolism, based on the
above definitions, provides a means of analysing the understanding of
mathematical forms, using the complementary concept.
• The ability to link form and content in mathematics is commonly considered a
main feature of what it means to understand mathematics (see e g Hiebert 1986).
The dynamic interplay between these four dimensions of mathematics and doing
mathematics (i e form and content, and structure and operation) gives a
characteristic of the logical and psychological aspects of mathematics (Bergsten
• Bearing in mind that all discussions of learning and understanding mathematics
in educational contexts by necessity must rest on some view of what
mathematics is, the conception of mathematics as a formalization of ideas
originated in our bodily experience or in other disciplines (like physics or
economics), as elaborated by Mac Lane (1986), puts an emphasis on
mathematical forms, as well as on its relation to the corresponding ideas and
activities, when it comes to understanding.

The notion of genetic forms is related to the idea of analogical reasoning (Gentner
1989, English 1997), in the algebra context discussed in English and Sharry (1996). It is
commonly argued that these ”surface” structures of mathematical symbolism are mere
conventions, and that the meaning they carry must be found by the individual’s own
mental constructions of their relationship to the ”deep” structure of mathematics, or the
algebraic manipulations become meaningless (English & Sharry 1996). Here it is
argued, however, that meaning often can be more or less directly evoked already at the
”surface” level by its very structure.

It should be observed that the notion of mathematical forms does not extend to a
possible more abstract notion of ”mathematical form”, in the sense of general structures
that are being formalized as mathematics. Such structures can to some extent be
expressed by non-algebraic forms of representation, such as verbal and pictorial forms.
The discussion in this paper is focussed on the implication for the understanding of
school algebra of the notions of mathematical forms and image schemata.
European Research in Mathematics Education I.II: Group 6 127

3. Mathematics and Image Schemata

The conceptual framework of image schemata by Johnson (1987) and Lakoff (1987)
has a tremendous potential impact on mathematics education. Its relevance to meaning
in mathematics and mathematical reasoning and understanding has been pointed out by
the authors themselves, and later, among others, by Presmeg (1991) and Dörfler (1991).

According to Johnson and Lakoff, meaning and understanding in human reasoning
are based on the use of image schemata. Our bodily experience of ourselves and our
physical environment, and our way of structuring and organizing this experience,
establish cognitive image schemata (with a non-propositional character) that depict
recurring regularities and structures of these experiences and organizing activities.
These schemata thus are constructed by the individual. Examples of some basic image
CENTER-PERIPHERY (see e g Johnson 1987, p. 21, 85, 113, 117, 121, and 124,
respectively). A key notion in the theory is that these kinds of bodily based schematic
structures also are used in human abstract thinking by the means of metaphorical
projection from the world of bodily experience into the abstract dimension.

In her study on visualization in mathematical thinking Presmeg (1985) found two
ways students used imagery to depict abstract situations, by concretizing the referent
(“making a concrete visual image the bearer of abstract information”), and by pattern
imagery (“embodies the essence of structure without detail”). The latter is a construct
similar to image schemata (Presmeg 1991, p. 7).

Dörfler (1991), stressing the holistic aspect of (subjective) meaning of mathematical
concepts, argues that the (mental) construction of appropriate image schemata
(referring to Johnson and Lakoff) is one means of facilitating the “cognitive
manipulation” of such concepts. This construction can be supported by working with
protocols of actions.

According to Mac Lane (1986) activities like counting, ordering, shaping, and
moving, give ideas of the common structures and regularities in these activities (across
situations), which can be formalized as mathematics. Thus the activity of counting
gives the idea of “the next one”, which can be formalized as “successor” by using the
European Research in Mathematics Education I.II: Group 6 128

Peano axioms (giving a formalization of the mathematical “object” ordinal number).
Behind mathematical concepts like limit and continuity is the idea of approximation,
originated in activities like estimating. More complex mathematical structures seem to
be generated from within mathematics itself. Thus it is appropriate to talk about
content-dominated mathematics and form-dominated mathematics (Bergsten 1990),
thereby indicating that the mathematical forms used in the former have a more direct
link to ideas based on bodily experience, or, in the above terminology, are
predominantly genetic.

Many mathematical notions can be related to ideas of an image-schematic character.
Set theory can be seen as the formalization of properties of the CONTAINER schema,
distance as the CENTER-PERIPHERY and SCALE schemata, a function as the LINK
schema and its graph as the PATH schema, and an equation as the BALANCE schema.
Many concepts and methods of mathematics seem to be developed from the human way
of thinking by the use of bodily based image schemata. Whatever the “real”
mechanisms behind the historical development of mathematical notions and notations
have been, the use of image schemata theory when discussing mathematics seems to
have a high potential value in the educational context. Dörfler writes (1991, p. 20):
“The cognitive manipulation of mathematical concepts is highly facilitated by the
mental construction and availability of adequate image schemata. In other words, the
subjective meaning of mathematical terms has a non-verbal, non-propositional and
geometric-objective component. The individual understanding of a mathematical topic
possibly is best grasped as a kind of interplay between the propositional expressions
and corresponding image schemata.” The last sentence from this quotation can possibly
be seen as an image-schematic interpretation of the above mentioned (common) view
that much of mathematical understanding relies on making connections between form
and content. The (subjective) meaning of mathematical content then is given by its
image schematic structure, a view in line with Mac Lane’s description of mathematics,
as indicated by Lakoff (1987, p. 365, referring to an article by Mac Lane from 1981):
“Mac Lane’s view of mathematics is thus very much like the view of human conceptual
systems that has emerged in this book.” The forms of mathematics “are those that
emerge from our bodily functioning in the world and which are used cognitively to
comprehend experience.” (Lakoff 1987, p. 365)
European Research in Mathematics Education I.II: Group 6 129

These ideas have recently been further developed by Lakoff and Núñes (1997), who
set up the new task of describing the ”metaphorical structure of mathematics” (p. 21),
and they write: ”One of the properties of commonplace conceptual metaphors is that
they preserve forms of inference by preserving image-schema structure.” (p. 30)

4. Mathematical Symbolism and Image Schemata

The relevance of the image schemata theory to (the understanding of) mathematical
symbolism has been noted by Dörfler (1991, p. 28): “I think that formulas in
mathematics can play the role of a carrier for an appropriate image schema. The latter
then is made up by the spatial relations of the symbols in the formula and of the
admissible operations and transformations with the formula. It is this image schema
which lends meaning to the formula as its concrete carrier”. Here similar structural
(figurative) and operational aspects of mathematical symbolism as those introduced in
Bergsten (1990), and discussed above (in the paragraph on mathematical forms), are

With this focus on the figurative aspects of mathematical symbolism, a key
observation is that genetic mathematical forms (as defined above) show the same
schematic structures as the BALANCE schema, the LINK schema and others, and this
almost by definition! That forms (schematic structures) are genetic means that they are
based on the way humans structure and organize their bodily experience. A possible
redefinition of the notion genetic mathematical form, keeping the intention behind the
original definition intact, relates such a form to a (similarly structured) prototypical
image schema behind the idea of the referent of the symbolic expression. Now, image
schemata are subjective, but this redefinition makes sense, since many basic schemata
are prototypical and socially shared.

When an idea is based on the activity of putting things together, it (by human
imagination) can be mentally structured by the LINK schema, by Johnson (1987, p.
113) schematically materialized by a picture like o-o . That the recording in some
notational form of such an activity or action (what Dörfler calls protocol of an action),
shows the same schematic structure as the image schema itself, is more likely to be
expected than not. The historical success of the mathematical form used in the standard
European Research in Mathematics Education I.II: Group 6 130

notation for addition, a+b, thus has a psychological explanation in its prototypical
image schematic character.

Johnson (1987, p. 90 and pp. 95-98) suggests that the BALANCE schema, with its
properties of symmetry, transitivity, and reflection, provides a possible experiential
basis for the ideas behind the mathematical concepts equality and equivalence relation.
In the schematic structure of the symbolism, the equality sign (in itself made up by two
equal parts, as stressed by Robert Recorde himself) serves as the fulcrum of the balance.
In the terminology above, the mathematical form of written equalities is a link (when
chunking of left and right “wings” is made), and must be classified as genetic by its
connection to the BALANCE schema.

Form-dominated mathematics naturally inherits some of its mathematical forms
from the content-dominated mathematics, making it possible to “understand” such
forms by metaphorical projection from the image schematic structures of genetic
forms. The notation for associativity of compositions of elements of groups, a purely
formal property, is (most likely) inherited from the notation for repeated addition (or
multiplication), exhibiting a genetic mathematical form. The ease of operating on such
purely formal expressions, as well as the sense of understanding (in some way) what
one is doing, can be explained by the image schemata that are being evoked by
observing the schematic structures of the formal expressions.

Dörfler’s (1991) recommendation to use protocols of actions (to promote the
construction of adequate image schemata) to develop understanding of mathematical
concepts genetically related to those actions, thus also applies to develop understanding
of mathematical forms. The common element is the construction of image schemata. In
the case of forms, however, the actions of interest are the reading, manipulation, and
interpretation, of (mathematical) symbolic expressions.

The notion of mathematical forms brings the focus to the structural (figurative)
aspects of (mathematical) symbolic expressions. Form operations deal with the
operative aspect of symbolic expressions, transforming them to new symbolic
expressions. Some common such operations have been identified above, and their
image schematic character is based on bodily experience. No doubt, activities like
splitting or joining (putting together), adding or deleting (taking away), and raising or
European Research in Mathematics Education I.II: Group 6 131

lowering, are prototypical across many experiential situations from low ages on. Also
the identity operation can be experienced, as when turning a cube upside down or
rotating a ball, leaving the object “unchanged”. It makes sense to refer to some form
operations as genetic, as splitting, joining, and mirroring, on some symbolic
expressions in arithmetic (see examples above), while others, like adding, deleting,
raising, and lowering (as exemplified above), are stipulated. Using Dörfler’s (1991)
terminology, image schemata related to mathematical forms are figurative, image
schemata related to form operations are operative.

5. Developmental Aspects of Understanding Forms

Real numbers are used (“applied”) in many different kinds of settings: though dealing
with a variety of situations (denoting magnitudes, proportions or time) they still obey
the same formal rules. To “understand” what real numbers are then by necessity
involves “understanding” the formal rules for such numbers. These formal rules are
“visible” in mathematical formulas such as a+b=b+a . This means that for a learner,
understanding of mathematical forms (in the sense defined above) is a path that
(possibly) can lead to an improved understanding of mathematics. Indeed, genetic
forms like in the above formula (for commutativity) have the potential of evoking the
same image schemata as the ideas or activities they depict. This way a link between
form and content can be (mentally) established, and the idea of understanding formulas
as “pictures” of image schemata can be evoked. To operate on formulas of genetic form
is like operating directly on the reality. They are isomorphic activities. This observation
may be one key to answering the classical question on why mathematics works, i e that
results of symbolic manipulations are applicable to reality (Bergsten 1990, p. 166; cf e g
Kline 1985).

Thus, in mathematical understanding there is a dynamic interplay between form and
content, facilitated by the use of image schemata. On the other hand, there is an
interplay between structural and operational aspects of form as well as of content. In her
analysis of conceptual development, Sfard (1991) stresses the complementarity of the
structural and operational aspects of mathematical concepts and entities.
Understanding the structural aspect of a concept presupposes the use of operational
European Research in Mathematics Education I.II: Group 6 132

aspects of concepts with a structure already understood. However, there is a vicious
circle: “the lower-level reification and the higher-level interiorization are prerequisite
for each other!” (Sfard 1991, p. 31). This can explain, Sfard argues, why for many
students the whole enterprise of mathematics can become a meaningless game of rules
producing correct answers. The reason is that reification has not occurred. Seeing a
structure is more difficult than performing an operation. Therefore many students need
well-planned educational activities for the reification process. The use of protocols of
actions to produce adequate image schemata (Dörfler 1991), has the potential of
facilitating this process.

Understanding mathematical forms seems to develop in a similar way. The
schematic structure of a formula can become visible by operating on instances where
the formula holds. For the “reification” of the form (of the formal notation) for
commutativity, for example, adding different pairs of numbers in both orders and
systematically recording these operations symbolically (e g in columns), the schematic
structure of these notations can be interiorized, performed and “seen” purely mentally,
and finally reified as a formal structure. In this case the known form used is the link
form of the addition formula, and the new form produced the form operation mirroring
for addition formulas. The sense of understanding these genetic forms is hypothesized
to be given by the similarly structured image schemata they evoke, originated in bodily
based activities and experiences.

Now, there is no need to understand the formula for commutativity (not to be mixed
with the idea that adding in reverse order yields the same result) until it must be used in
a formal computation. In school this normally happens only after algebraic notation has
been introduced, and that is also when understanding mathematical forms is beginning
to become crucial also for mathematical achievement. For some students the algebraic
terms themselves are void of meaning: they are simply symbols (atomic forms) like a,
b, and c, in written (rule-governed) expressions like a+b=c. Thus in beginning algebra
there is a risque for a double meaninglessness: operations are performed on algebraic
expressions, made up by terms without meaning, and with a form without meaning.

To develop an understanding of mathematical forms, and pave the way for symbol
sense, its structural and operational aspects must both be grasped. These seem to
presuppose each other, i e are complementary for this understanding. To apply the form
European Research in Mathematics Education I.II: Group 6 133

operation joining on the left wing of the expression 7x-x=54, to obtain 6x=54, the
schematic structure of the expression must be understood (possibly by the LINK and/or
BALANCE image schemata). On the other hand, to see the basic structure, the link
around the equality sign, it must be observed that a joining operation can be performed
on 7x-x.

6. Symbol Sense

Some aspects of symbol sense in the context of school algebra have been described by
Arcavi (1994), one of which is strongly related to the present discussion: ”[What is
symbol sense?] - An ability to manipulate and to ”read” symbolic expressions as two
complementary aspects of solving algebraic problems.” (p. 31) It is argued in this paper
that the development of this (key) aspect of symbol sense can be facilitated by building
on the recognition of the role of image schemata and genetic mathematical forms for the
feeling of understanding the mappings between form and content in school algebra.
These constructs provide a link from sense impressions to the development of symbol
sense in elementary school algebra.

Higher-level symbol use in (form-dominated) mathematics rests heavily on
familiarity with the basic algebraic language used in school. The role of
non-propositional, imagistic thinking is furthermore prevalent at all levels of
mathematical thinking (cf Sfard 1994).

7. References
Arcavi, A. (1994): Symbol sense: Informal sense-making in formal mathematics. For the Learning of
Mathematics 14 (1), 24-35.
Bergsten, C. (1990): Mathematical operativity. An analysis of the form-content relationship in school
mathematics. (In Swedish, with an English summary) Linköping Studies in Education.
Dissertations No 29. Linköping: Linköping University.
Cajori, F. (1929): A history of mathematical notations, Volume II: Notations mainly in higher
mathematics. Chicago: Open Court.
European Research in Mathematics Education I.II: Group 6 134

Dörfler, W. (1991): Meaning: Image schemata and protocols. In F. Furinghetti (Ed) Proceedings
Fifteenth PME Conference, Vol. I, 17-32. Dipartimento di Matematica dell’Universita di
Genova, Italy.
English, L. (Ed.) (1997): Mathematical reasoning. Analogies, metaphors, and images. Mahwah, NJ:
Lawrence Erlbaum Associates.
English, L. & Sharry, P. (1996): Analogical reasoning and the development of algebraic abstraction.
Educational Studies in Mathematics, 30, 135-157.
Gentner, D. (1989): The mechanism of analogical learning. In S. Vosniadou & A. Ortony (Eds.),
Similarity and analogical reasoning. Cambridge: Cambridge University Press.
Hiebert, J. (Ed) (1986): Conceptual and procedural knowledge: The case of mathematics. Hillsdale,
NJ: Lawrence Erlbaum Associates.
Johnson, M. (1987): The body in the mind. The University of Chicago Press, Chicago.
Kline, M. (1985): Mathematics and the search for knowledge. New York: Oxford University Press.
Lakoff, G. (1987): Women, fire and dangerous things. The University of Chicago Press, Chicago.
Lakoff, G. & Núñes, R. (1997): The metaphorical structure of mathematics: Sketching out cognitive
foundations for a mind-based mathematics. In L. English (Ed.). Mathematical reasoning.
Analogies, metaphors, and images.
Mac Lane, S. (1986): Mathematics, form and function. Springer-Verlag, New York.
Picciotto, H. & Wah, A. (1993): A new algebra: Tools, themes, concepts. Journal of Mathematical
Behaviour, 12, 19-42.
Pimm, D. (1995): Symbols and meanings in school mathematics. London: Routledge.
Presmeg, N.C. (1985): The role of visually mediated processes in high school mathematics: A
classroom investigation. Unpublished Ph.D. dissertation, University of Cambridge.
Presmeg, N. C. (1991): Metaphors, metonymies and teacher beliefs about visual processing in
mathematics classrooms. Paper presented at the Fifth International Conference on Theory in
Mathematics Education, Paderno del Grappa, June 20-27, 1991.
Sfard, A. (1991): On the dual nature of mathematical conceptions: Reflections on processes and
objects as different sides of the same coin. Educational Studies in Mathematics, 22 (1), 1-36.
Sfard, A. (1994): Reification as the birth of metaphor. For the Learning of Mathematics, 14, 44-55.
Sierpinska, A. (1995): Understanding in mathematics. London: Falmer Press.
European Research in Mathematics Education I.II: Group 6 135

Pilar Bolea, Marianna Bosch, Josep Gascón
Departament de Matemàtiques – Edifici C, Universitat Autònoma de Barcelona, 08193 Barcelona, Spain

Abstract: Algebra is considered here, not as a particular mathematical organization
(such as arithmetic or geometry, for instance) but as a process, the process of
algebraization, that can affect either a whole mathematical organization or, as is
nowadays the case in secondary education, some aspects of it. Depending on the case
examined, a more or less algebraized mathematical work results: the algebraization
process can in fact be considered as the modelling of a whole initial mathematical work
in order to deepen our knowledge of it. Algebraization is thus, in this sense, a
specifically didactic activity, relative to the study of mathematics. Institutional
restrictions on the algebraization process constitute a paradigmatic example of the fact
that mathematics (the object of study) and didactics (the process of study) are
Keywords: algebraization, process of study, mathematic organisation.

1. Mathematics as the Study of Mathematical Works

In their daily work as teachers, students and researchers, mathematicians usually
describe mathematical knowledge in terms of concepts, notions, intuitions, ideas,
methods, processes, definitions, problems, etc. They can then talk about ”students’
difficulties in the acquisition of a concept,” or about the ”inability to apply a notion
correctly in order to develop a new method,” ”the importance of being able to formulate
a good definition,” ”the danger of focussing on processes while disregarding the
general idea hidden behind them,” etc. What is used, in all these cases, is the common
epistemological model of mathematics, that is, the usual way of perceiving,
interpreting, describing, and thinking about what mathematics is and what
mathematical activity exactly consists of.
European Research in Mathematics Education I.II: Group 6 136

The anthropological approach to the didactics of mathematics (Chevallard 1992)
establishes that institutional mathematical practices are didactics’ primary object of
study, which means that, in an initial research stage, mathematical activities cannot be
taken for granted, nor considered as a primitive concept but, on the contrary, as entities
that, just like any other object of scientific study, need to be questioned and modelled.
In order to achieve this, the anthropological approach proposes a general
epistemological model that describes mathematical knowledge in terms of
mathematical works or praxeologies, and presupposes that these mathematical works
are structured into institutional mathematical organizations, which, in turn, are made up
of four main components: several types of (problematic) tasks, techniques (to carry out
these tasks), technologies and theories. The pair [task/technique] constitutes the praxis
(or ”know-how”), whereas the pair [technology/theory] is the logos (or ”knowledge”)
of the whole praxeology, which includes both praxis and logos. This theoretical
framework also describes the process of creation of these four components, that is, the
way in which a technique can emerge from the study of a problematic task (or
”problem” for short), its development into new techniques, the way in which certain
descriptions and justifications of this work become a technology (i.e., a discourse
—logos— about a technique —technè—), etc. (Chevallard, Bosch & Gascón 1997;
Chevallard 1997).

Generally speaking, mathematical activity can be considered as the use of a
mathematical organization or a mathematical work. But it is also, at the same time, a
production (or re-production) of mathematical realities that will lead to new
mathematical organizations. The English term ”work” (translated form the French
œuvre) allows us to talk about mathematics as a human activity —given that
mathematics are something we do— and as an artifact produced and reproduced by this
activity —the work of mathematicians—. A mathematical work is something to be used
and something to be produced or reproduced. Taking the word ”study” in its broadest
sense, the anthropological approach considers mathematical activities as processes of
study of works within given institutions. This process of study —or didactic process—
includes mathematical research, which always begins with the study of open questions
and problems, as well as the so-called teaching and learning process. The didactic
process takes place in a community (groups of researchers, of students, etc.), is usually
supervised by a director of study (be it the main researcher, the teacher, an advanced
European Research in Mathematics Education I.II: Group 6 137

student, etc.), and has its own program of study (an open problem, a whole research
program, the curriculum, etc.).

According to the four-component structure of mathematical works, the process of
study can be described in terms of six didactic moments or dimensions: the moment of
the first encounter, the exploratory moment, the moment of the technique, the
technological-theoretical moment, the institutionalization moment, and the evaluation
moment. It is important to note in this respect that we speak about ‘moments’ in a
functional, not chronological, way, as when we say that something is bound to happen
or that ”there is always a moment when ...”.

Within this theoretical framework, we would like to emphasize a rather surprising
aspect of the link existing between didactics and mathematics. We start by considering
the algebra related to the algebraization process of a mathematical organization. This
process can be defined as a process of mathematization which consists in the
construction of a new mathematical organization that models the given one. We
therefore do not consider school algebra as a mathematical organization in itself, but as
a way of modelling a given mathematical organization so as to make its study easier.
We can talk, in this sense, of a didactic technique, that is, a special way of studying a
mathematical organization. This point of view allows us to analyse the consequences
that the algebraization process has on the study of a mathematical organization and to
stress a number of institutional restrictions imposed on the process of study of
mathematics in secondary education.

2. The Need of a Model of School Algebra

In Gascón (1994) we analysed the didactic phenomenon of school arithmetization of
elementary algebra and related this phenomenon to the interpretation —which is rather
common in pre-university academic institutions— of algebra as 'generalized
arithmetic' (in the sense expounded in Booth 1984, Drouhard 1992, Filloy & Rojano
1989, Kaput 1996, Vergnaud 1988). We also maintained that it is necessary to go
beyond the mere questioning of this interpretation to take it as an object of study, i.e. as
an empirical fact that didactics should be able to explain. But this can only happen if, on
the one hand, we establish a notion of algebra that will allow us to interpret 'the study of
European Research in Mathematics Education I.II: Group 6 138

algebra' in a given institution, and, on the other hand, we use it as the basis to generate a
series of didactic phenomena related to what is commonly called ”the learning process
of algebra.” Contrary to any given naturalistic point of view, these phenomena, just like
psychological, sociological, historical or physical phenomena, have to be constructed
scientifically and cannot be taken for granted.

Elementary algebra does not appear as a self-contained mathematical work
comparable to other works studied in academic core courses (such as arithmetic,
geometry, statistics, etc.), but rather as a modelling tool to be (potentially) used in all
mathematical curricular works and which appears to be more or less used in them.
According to that, the model of elementary algebra that we have chosen as an
alternative to 'generalized arithmetic' is based on the realization that elementary algebra
is in fact a mathematical tool, the algebraic tool, that can be used to study many
different kinds of problems not only or exclusively pertaining to arithmetic, which
leads to the notion of algebraic modelling as described by Chevallard (1989). We will
therefore not talk here about the teaching of 'algebra' as we do about 'arithmetic' or
'geometry'. We will consider school algebra, not as an organization that can be studied
per se, but rather as a mathematics set-up that makes the existence of a given process
—the process of algebraization— possible, and which does not affect the mathematical
works studied at school always in the same way.

Since mathematical activities (including school mathematics) appear, from a given
stage of development, as being fully algebraized, in the sense that they cannot be
conceived of without the whole functionality ascribed to the algebraic tool, we have to
admit the existence of a process of algebraization of school mathematics. This process
starts with primary education, continues throughout secondary education and is
completed at the university level. Although this process may take place, on certain
occasions, in an explicit way that can be recognized by the actors of the institution, it
emerges more often than not in a subtle, surreptitious way, depending on the
mathematical needs arising at any given moment, the actors of the institution usually
becoming aware of its existence without even having witnessed its actual emergence. In
secondary education, mathematics tends to have a clear 'pre-algebraic' nature, as if it
remained foreign to the algebraization process and would not even allow this process
—which happens to be unavoidable— to take place in a coherent and controlled way, in
accordance with its corresponding level (Gascon 1998).
European Research in Mathematics Education I.II: Group 6 139

Our main purpose is to start describing, analysing and characterising the multiple
links existing between the degree of algebraization of a mathematical organization and
the possible ways of conducting its study. We shall also show, reciprocally, that the
different kinds of constraints of a mathematical, didactical and cultural nature which
are imposed on the process of algebraization of mathematical works at school
constitute, in turn, restrictions to the didactic process, that is, to the specific way in
which these works can be studied at school.

3. The Algebraization of a Mathematical Work

It is important to point out that we do not have any demarcation criteria to delimit
precisely what an algebraized work is and differentiate it from a pre-algebraized one (in
the sense of ”not yet algebraized”). We thus postulate that there is always a question of
degree in the algebraization of a mathematical work, and we will try to characterize this
process by means of the notion of modelling (Chevallard 1989). More specifically, we
will state that a mathematical work is algebraized if it can be considered as an algebraic
model of another mathematical work, the system to be modelled.

Such a definition clearly transfers the problem in question to the characterization of
algebraic modellings. Although we cannot offer a complete description of them here,
we can at least present some of their characteristic features, which will allow us to
distinguish algebraic modellings from other kinds of mathematical modellings:

(3.1) The algebraic modelling of a given mathematical work describes explicitly and
materially all the techniques contained in the initial work, thus allowing for a quick
development of these techniques, as well as for the explicitation of their interrelations
and the unification of the related types of problems.

(3.2) An algebraic modelling can be considered, in this sense, as the answer to a
technological questioning related to the initial work, such as, for instance, the way in
which to describe and justificate the initial techniques, the conditions under which they
can be applied, the types of problems they can solve, etc.
European Research in Mathematics Education I.II: Group 6 140

(3.3) In an algebraic modelling, all components of the initial work are modelled as a
whole, and not as separate entities, a fact which tends to simplify the structure of the
algebraized work eventually obtained.

EXAMPLE: Let us consider, as pre-algebraic work to be modelled, the usual
mathematical organization built up around elementary divisibility problems, such as:
how to find the multiples and divisors of a given number, how to set up divisibility
criteria, how to calculate common multiples and divisors of two or three numbers, the
greatest common divisor or least common multiple of two or three numbers, the list of
the prime numbers down to 100, etc. The associated techniques are based on the
fundamental properties of multiplication and division of integers, which make up the
core of the technological-theoretical component. These techniques take the form of
numerical tables which contain the multiples of two numbers or the multiples of the
smaller number and the difference between the first two, the divisors of two numbers,
the successive differences between two numbers and between the smaller one and the
difference obtained, etc. (see below) and are bound to be developed until the standard
form of Euclid’s algorithm is obtained.


240 60 (=300-240) 300 551 437 551-437=114

480 120 114 323 323-114=209

720 180 114 209 209-114=95

960 240* (rep.) 1200 114 95 114-95=19

1200 300 19 95 95=19ÿ5

1440 360 19 0 19 = (551,437)
... ... ...
European Research in Mathematics Education I.II: Group 6 141

The technological questioning of such a work affects the description, justification
and scope of the techniques in use. One may begin, for instance, with the following
questions: Why does a common multiple of two numbers turn out to be a multiple of
their difference? Why is the first repeated multiple in the table the least common
multiple? Why calculate differences when looking for a common divisor? How does
this relate to Euclid’s algorithm? How can we interpret partial results of this algorithm?
In which cases can we simplify it? etc. Algebraic modelling of the arithmetical
techniques based on numerical tables should lead to the consideration of linear systems
of Diophantine equations. Thus, the above tabular technique for the determination of
the l.c.m. of 240 and 300 can be modelled with the linear Diophantine equation 240x +
(300 - 240)x = 300x, which determines a common multiple of 240 and 300 if we can
find an integer y such that (300 - 240)x = 240y, that is 60x = 240y. The solution x = 4 and
y = 1 gives the l.c.m., i.e. 240·(4 + 1) = 300·4. This kind of modelling allows the
description of the original types of problems and the explicitation of possible links
between them (specially between the l.c.m. and the g.c.d. of two numbers). It can also
answer the technological question mentioned above and contribute to the formulation
of new types of problems (for instance, which conditions have to verify two numbers a
and b if they have 24800 as l.c.m.?, which common divisor of 120 and 200 is a multiple
of 6?, etc.). Finally, the model based on linear Diophantine equations both unifies and
simplifies the structure of the algebraized organization, and it also produces a clear
ostensive reduction of the scripts used.

4. Characteristics of an Algebraized Mathematical

Having already described an algebraized mathematical work as being the result of some
algebraic modelling, we shall now present a few indices of the degree of algebraization
of a given mathematical work:

(4.1) In algebraized works, modelled techniques take place at a technological level
with regard to the original pre-algebraized work. An index of algebraization of a given
mathematical work is therefore linked to the possibility of formulating and studying
technological questions related to the description, interpretation, justification and
European Research in Mathematics Education I.II: Group 6 142

validation of the initial work.

(4.2) In particular, the more a mathematical work is algebraized, the more it enables
us to describe the different types of problems that can be solved, as well as the
necessary conditions for solutions to exist, their possible uniqueness and their
structure. In an algebraized organization, the main type of problems is no longer the
determination of solutions but the study of their conditions of existence. Consequently,
problems will not be studied in themselves, as isolated entities, but as components of a
given type of problems.

(4.3) A mathematical work is constructed to answer some type of problematic
questions that, in turn, give rise to various types of problems. Thus, an indicator of the
algebraization degree of a given work is linked to the possibility of considering,
describing and handling the global structure of the above-mentioned problems.

(4.4) In an algebraized work, we use parameters and variables systematically, both
in mathematical techniques and in the associated theory. This use includes the
manipulation of formulas and is supposed to give rise to a functional language.

(4.5) An algebraized mathematical can easily become independent from the original
system that it originally models. More specifically, the possibility of generating
problems that are detached from their context of emergence constitutes an indicator of
the degree of algebraization of a given work.

(4.6) The algebraization of a work is also related to the unifying of the different types
of problems that are contained in this work, as well as to the integration of its
corresponding techniques and technological elements.

(4.7) This simplification of the components of a given algebraized work leads to a
considerable reduction of the 'ostensive' material (such as words, writings, graphics,
gestures, etc.) that is used to develop mathematical activity. In other words, an
algebraized mathematical activity needs less 'ostensive material' to be worked out, but a
material which is much more instrumental (Bosch, 1994).
European Research in Mathematics Education I.II: Group 6 143

5. The Role of Algebraization in the Study of a Mathematical

It is essential, in this context, to try and analyse the actual or potential consequences of
the algebraization of a given mathematical work as regards the many possible ways in
which this process can be conducted in modern educational institutions. These
consequences need to be confronted to the institutional didactic contract that currently
governs the study of mathematics at school. We will be using here, to this effect, the six
dimensions of the process of study involved in the anthropological approach to

(5.1) In the process of study of a given mathematical algebraized work, the
exploratory moment has a necessarily 'material' nature. Indeed, exploration is not
carried out thanks to a purely 'mental' process, but rather with the crucial participation
of handwriting. In other words, most of the ”plausible reasoning” (in the sense of Pólya)
is carried out through calculation. More than any other, the algebraization process
reminds us that mathematics is a 'material' activity, that it cannot be achieved without
resorting to material instruments (be them oral, written, graphic, or ‘gestural’). This
characteristic aspect of algebraized mathematical activity has to be contrasted with the
currently existing cultural illusion —reinforced by the dominance of pre-algebraic
”mathematical thinking”— that mathematical activity takes place ”in our mind” only.

(5.2) If we want students to be able to develop by themselves the techniques they
use, instead of having these techniques presented to them by the teacher, then the
process of study of a mathematical organization is to go beyond the exploratory
moment and the routine application of ready-made techniques. A 'material'
representation of these techniques will then be needed to permit their explicitation and
to start the technological questioning about their scope, interpretation, and
justification. We can then say than the algebraization process not only emerges from a
technological questioning about a previous given work, but that it also allows this kind
of questioning. In other words, the technological questioning about the description,
interpretation and validation of a given work appears closely and doubly linked to the
process of algebraization of this work. Difficulties about bringing into being the
technological-theoretical moment are coherent with the pre-algebraic nature of the
European Research in Mathematics Education I.II: Group 6 144

mathematical works studied in secondary education, and, consequently, with the
obstacles to the complete algebraization of these mathematical works.

(5.3) In the process of study of a given algebraized work, the above-mentioned
technological questioning leads to the establishment of fluid interrelations between the
moment of the technique and the technological-theoretical moment. The flexibility
characterizing the techniques used in this process is thus reinforced, and so is the power
inherent in the moment of the technique to integrate the different components of the
mathematical organization in question. In the study of a pre-algebraic mathematical
organization, such relations are more difficult to establish, which produces a certain
rigidity in the use of techniques and impedes the study of the mathematical organization
as a whole.

(5.4) Pre-algebraic mathematical activity always implies a certain 'atomization' of
its components due to the fact that the cultural interpretation of 'concrete' systems, as
well as the apparent 'naturalness' of pre-algebraic techniques (especially arithmetical
and geometric ones), provide in advance the necessary intelligibility for its survival in
the institution where this activity is bound to take place —even though, in most cases,
this intelligibility might turn out to be purely 'local'—. On the other hand, it is to be
noted that any algebraized organization is bound to lose a considerable amount of
cultural intelligibility, while preventing the activity in question from being atomized,
being as it is the result of a sustained and continuous process of study that is in constant
need of interpreting and evaluating the original organization globally. Again, the
pre-algebraic nature of mathematical organizations studied at school is coherent with
the absence of medium-term and long-term didactic goals, as well as with the
difficulties inherent in the didactic process as regards the institutionalization and
evaluation of a whole mathematical organization.

We have emphasized the strong relationship between the structure of the
mathematical organizations liable to live in a given didactical institution (such as
secondary education) and the kind of didactic processes that can be made to exist in that
institution. More specifically, we have shown that the algebraization of a mathematical
work can be considered as a didactic tool that facilitates the process of study of a
mathematical organization as a whole, and leads to what we will call the integrated
study of a mathematical work. Starting from this point, many questions should be
European Research in Mathematics Education I.II: Group 6 145

raised, especially those relating to the conditions under which the ”algebraic-didactic”
technique can be used in the classroom. What are the most appropriate pre-algebraic
mathematical organizations to be modelled? What kind of institutionalization should
they undergo to ease their algebraization? How to induce a technological questioning
about these organizations? At what level can we consider the explicit algebraic
modelling of mathematical techniques as an object of study? In short, to what extend
does the kind of didactic process conducted in secondary education reinforce or, on the
contrary, prevent the full algebraization of mathematical works?

6. References
Booth, L. (1984): Algebra: Children’s Strategies and Errors. Windsor: NFER-Nelson.
Bosch, M. (1994): La dimensión ostensiva en la actividad matemática. El caso de la
proporcionalidad. Doctoral dissertation, Universitat Autònoma de Barcelona.
Chevallard, Y. (1989): Arithmétique, algèbre, modélisation. Étapes d’une recherche. Publications de
l’IREM d’Aix-Marseille, 8.
Chevallard, Y. (1992): Fundamental concepts in didactics: perspectives provided by an
anthropological approach. Recherches en Didactique des Mathématiques - Selected papers,
Chevallard, Y. (1997): Familière et problématique, la figure du professeur. Recherches en Didactique
des Mathématiques, 17/3, 17-54.
Chevallard, Y. , Y., Bosch, M., Gascon, J. (1997): Estudiar matemáticas. El eslabón perdido entre la
enseñanza y el aprendizaje. Barcelona: ICE/Horsori.
Drouhard, J.-P. (1992): Les écritures symboliques de l’algèbre élémentaire. Doctoral dissertation,
Université Paris 7.
Filloy, E., Rojano, T. (1989): Solving Equations: The Transition from Arithmetic to Algebra. For the
Learning of Mathematics, 9.2, 19-25.
Gascon, J. (1994): Un nouveau modèle de l’algèbre élémentaire comme alternative à l’”arithmétique
généralisée”. Petit x, 37, 43-63.
Gascon, J. (1998): La naturaleza prealgebraica de la matemática escolar. Educación matemática.
Kaput, J. (1996): ¿Una línea de investigación que sustente la reforma del álgebra? (I) & (II). UNO, nº
9, 85-97, nº 10, 89-103.
Vergnaud, G. (1988): Long terme et court terme dans l’apprentissage de l’algèbre. In Laborde, C.
(eds): Actes du premier colloque franco-allemand de didactique, 189-199. Grenoble: La Pensée
European Research in Mathematics Education I.II: Group 6 146

Veronica Hoyos, Bernard Capponi, B. Geneves
Univ. Pedagogica Nal. (Mexique) & EIAH-Lab.Leibniz/IMAG (France), 46, avenue Felix Viallet,
38000 Grenoble, France

Abstract: This paper focus on the connection between geometrical constructions and
algebraic descriptions derived of simulate some mathematical machines in CABRI-II
learning environment. We present a scenario experimented with pupils in a first degree
of a high school (eleventh grade). The action’s motive is modelling a Descartes’
machine as a CABRI-II diagram. The starting hypothesis leads on the fact that such
situations promote connections between geometrical properties and their dual
algebraic symbolisation. The animation of CABRI-II diagrams provides the
visualization of geometrical aspects, and calculus on geometrical objects provides the
algebraic symbolisation. This inquire into the meanings emerged of this mathematical
experience can be completed by asking the pupils how to refute (using the geometrical
tools and objets experienced) a frequent statement about the usual algebraic inequality
at this academic level.
Keywords: algebra, dynamic geometry, microworld teaching.

1. Introduction

About the signification of algebraic variables
In a previous study (Hoyos 1996 and 1998), we have identified several significations
that high school students (in Mexico from 16 to 18 year old) attach to linear equation
after having followed a traditional education course of analytic geometry (according to
European Research in Mathematics Education I.II: Group 6 147

a text like Fullers’ 1979). We have documented1 with the case study the importance that
the pupils attach to the operational character of the algebraic equations, the one which
expresses the form ax + by = c.

Besides, other authors (Schoenfeld et al., 1993; Duval, 1988; Herscovics 1980) had
already emphasized on the difficulties the students have when they are confronted with
establishing links between the graphical and the algebraic representations, especially
for the equations lineaires with two unknown. In the work we are presenting here, we
propose simulation’s scenarios on CABRI-II which are seeking for the links’
establishment between the usual graphic and algebraic representations in high-school
(eleventh grade). In the referenced studies (Schoenfeld et al. with the implementation to
the software GRAPHER, and Duval with the proposal emerged of didactical analysis)
these authors have also looked for the establishment of links between the graphic and
the algebraic representations. In this pursuit, we try to contribute to the consideration of
the algebraic equation’s emergence on an epistemological point of vue, like the
dynamic plot of curves and the Thales theorem’s application.

2. Class Room Context
This presentation shows a short part of a function’s teaching process in a eleventh grade
school class in France. The all process can’t be described here but we can give nine
main points studied in the class room.
1. Variable notion ( dynamic variable visualisation with counter and axis in Cabri II)
2. Functional dependance and proportionality
3. Different frames (geometry, mechanics..)
4. Graphic construction and interpretation related to geometrical problems in dynamic
5. Graphic frame ( graphic’s geometry)
6. Equations and inequations solving with graphic method.
7. Use of Algebra in analytical geometry (Descartes’ Machine, Circle equations).
8. Study of classical functions ( French curriculum)
9. Evaluations.
European Research in Mathematics Education I.II: Group 6 148

In all the process students use classic methods with paper and pencil but also a part
important of graphics with graphic calculators and with the Cabri-geomètre II
software2 . The process was used in classroom from September 1997 to June 19983. All
students knows the software Cabri-geomètre by using it in the mathematical course for
geometrical studies since the beginning of year.

3. Theoretical Frame of our Educational Experience

Descartes’ voice in the classroom
We present here the introduction of the simulation on CABRI-II of Descartes’ machine4.

This machine, to plot hyperbolas, is made of a mechanism, composed of a system
(RSQ) of a fixed slope that slides vertically and of a shaft (AF) tied to this system. This
shaft (AF) is also articulated to a fixed point (F) on a horizontal axis (see fig.1). By
introducing the simulation on CABRI-II of the mechanical construction of the plots, we
wish to get close to apply the fundamental elements of Descartes’ analysis situations in
order to obtain algebraic equations.





Fig. 1

Our teaching proposal has been complementary of the normal mathematics course
of the first year5 of high-school (eleventh grade class). We set two practical work (PW)
up, one of which (the first) was constituted by three sessions of practical work of one
hour each.
European Research in Mathematics Education I.II: Group 6 149

The task’s sequence of this first PW has been:
1) Modelling the machine.
2) Algebraic characterization of the straight line’s families that step in the plot of
Descartes’ hyperbolas
3) Obtaining the equations of hyperbolas.

We complete this instructional work with other activities (statement of the theme
and graphic representation of the functions) organised around the refutation of one of
the solution’s procedure ( specially the crossed product and the table sign’s change) of
algebraic inequality, usual for the pupils and yet incorrect. To resume, we have
introduced a « Descartes’ voice » in the classroom of eleventh graders, looking for a
production of « echos6 » that could lead to the establishment of geometrico-algebraical
links. All the produced voices would give resources to the pupils to validate algebraic
executions usual at this academic level.

4. The Sources and the Techniques

Descartes’ machine
In the book II of La Geometrie , Descartes presents a method of curves’ generation
according to which we can obtain a curve’s family to a given algebraic curve.

His procedure was as follows: we start from a previously constructed curve SQ, S a
fixed point on the curve and A a point that is not on the curve; both points are fixed
regarding to the given curve. Be O, a fixed point on the straight line RS, and F, a fixed
point on the line perpendicular at the straight line RS in point O. Be P, an intersection
point of the curve with the straight line FA. So, when the curve QS (also A) moves with
a rigid translation movement parallel with RO, the point P will plot a new curve PF,
which can be considered as the daughter of the original curve. Thus, if the given curve
SQ is a straight line, the new curve will be an hyperbola. Here (fig 1), we can see a
simulation of this machine on CABRI-II.

Mathematical experience in CABRI-II
We started by showing a video (5 mn) which we especially elaborated (and which is
European Research in Mathematics Education I.II: Group 6 150

available from the EIAH7 team) to introduce the theme of drawing machines. In this
video, we talk especially about mathematical experience, specifically practice or the
use of tools to do mathematics, and we show the advantages of the simulation on
CABRI-II of this kind of machines. We asked then pupils to modele in CABRI-II
Descartes’ machine, after seeing it in the video.

Indeed, the students (working in CABRI-II) committed them selves in the proposed

The sessions of first practical work (PW), and the results:
1) The first session of this PW aimed to the simulation on CABRI-II of Descartes’
machine, the one the video just showed. For this construction, we had given the
following instructions with figure 1 :

“Descartes’ machine is constituted of two perpendicular lines and of a
right-angled triangle that slides on the vertical line when we move the point S.
During the moving, the triangle does not deform. The points R and Q cannot be
directly moved. A is a point of the segment RS which position can be modify.
The point F is a point of the horizontal line we can move. The length of the [DE]
can be modified and allows to adjust the length of [RQ]. Once the machine is
made, use the tool trace for the point P, and drag S to see the curve plotted by
the machine. Save your machine in a file.

The students were very interested in the task, and half of the pupils succeeded in the
modelling during the planned time session (approximately 50 mn). Between the pupils’
approaches to note, there is the one to make translation tool of CABRI-II like a way to
drag an object with construction’s constraint, like the orthogonality of the sides.

2) For the second session, we asked the pupils to characterize, using their algebraic
knowledge, the families of lines SQ (Fig 1) which intersections with AF give the plot of
the curve, the one the students had observed at the beginning of the first session. We
then asked the following questions :
European Research in Mathematics Education I.II: Group 6 151

1. Put an orthonormal landmark on the cabri-screen from which you have
made your machine. Draw the straight line SQ. Display the coordinates of
the points S and Q. By moving S on the vertical axis, we obtain a family of
straight lines (SQ). Put S in different positions (6 for example) and write
down on your sheet each time the coordinates of S8 and Q corresponding
to a each position. Then, construct the different straight lines by placing
the points noted on the grill. Do the same work on the drawing sheet that
has been given to you.
2. Do the straight lines you have drawn have something in common? and
what? Name d1,d2,d3,d4,d5,d6 the six straight lines.
3. Choose a pair of points (S,Q) among the one you have noted and determine
the slope of the corresponding straight line(SQ).
4. Redo 5 times the same thing with the other pairs of points
d2 slope:
d3 slope:
d4 slope:
d5 slope:
d6 slope:
So, in your opinion, the straight lines d1, ..., d6 are a family (yes, no, why).
5. We now take an interest in the straight lines (AF) when S moves along the
vertical axis. Are they also forming a straight line’s family (yes, no; if yes:
6. Use your modelling of Descartes’ machine to plot the curve with locus tool
Place 3 points on that curve and display their coordinates. Check with
calculation that these points are not aligned.

Pupils have been very interested during this second session of the PW too, and it
turned out to be a good pinpointing context for revisiting the them of the slope of the
straight line.

3) Finally, the third session of the first PW, the one that aimed to find the equation of
Descartes’ hyperbola, turned out to be the toughest. We had top give precise indications
of where to apply Thales theorem. See for example, the following part of the PW that
pupils have worked:

4. Now, it is about applying Thales to the triangles ASN and this one formed
by the vertex S and the measure’s segment x, which we are going to name
European Research in Mathematics Education I.II: Group 6 152

a) Obtain a cabri-construction like the one as follows:
b) Write here a proportionality relation between the perpendicular sides of
triangles ASN and MSP indicated.
The relation found is: ________________
c) Knowing that the straight line’s slope determined by SQ is fixed and
known, give him the value 1. You have to find an expression for the
segment MS in terms of x and of this slope’s value.
MS= ______________
5. Knowing that the points A and F are also fixed points, known points, we
are going to place them as follows:
- place the point A at a distance of 1 from point S
- place the point F on (1;0)
6. Because the value of AS =1, we have to express the segment AM in terms of
this value 1 and of x.

AM= _____________
7. Now, write here a proportionality relation between the indicated
perpendicular sides of the triangles AOF and AMP. The relation found is:





Fig. 2
European Research in Mathematics Education I.II: Group 6 153

Several pupils were close to obtain the equation of the curve. During the timed
session, approximately 50 mn, despite its interest and the planned guidance, the
students have only been able to establish the proportionality relations asked for. It was
outside the class, later, that we asked advanced pupils to finish that task. After
reworking ten minutes, they wrote the equation in question: ( f(1;x) - 2 + x = y. We
have completed this teaching experience with another second PW of two sessions of
one hour each, about the graphical representation of basical functions at CABRI-II, in
relationship with solving the algebraic inequalities (like f((x-1)2;x) = f(x2;x+1) and
others basical inequalities for this academic level).



4,70 x

Fig. 3

This kind of inequation can’t be solved by mean of table process because when
solving you need the study of the sign of - x2 - x + 1 . At this level students have not tools
to solve it. So they need to use a graphical method with a graphic calculator or with
Cabri-geomètre II.
European Research in Mathematics Education I.II: Group 6 154

First, we have noticed that most of the students were using the strategy of the
crossed product, or that the table sign’s change, to solve them. In this first case the
pupils would solve without being preoccupied by the corresponding signs. In the
second case the pupils couldn’t find the correct solution. Already, at the totality of the
second PW, we ask the pupils for the graphical pinpointing in CABRI-II of the curves
denoted by each member of the inequalities. Finally, we have got one hour session of
« echos » production, specially for reassess the connexions established about the
geometrico-algebraic procedures studied, by argumenting with the pupils about their
solution’s procedure for the inequalities; inequalities similar to these they have already
been confronted with (in the assessment sessions). They will also be able to evaluate
their disposal tools worked in CABRI-II.

We think that this revision of resolution’s own procedures and construction of
controlled tools give the pupils the occasion to validate their usual algebraic
manipulation as long as an analytical solution to the kind of exercises demands the plot
of hyperbolas, straight lines, and parabolas; graphical drawings which plot and
equation will be called by the introduced “voices”.

In conclusion, the work hypothesis that we assume here is that a structuring of work
sequences like these we presented would allow the pupils to establish links between the
graphical drawing of the curves and their corresponding algebraic expressions;
connexions that could more particularly provide the mathematical experience
developed in9 CABRI-II.

5. References
Boero, P., Pedemonte., B., Robotti, E. (1997): ‘Approaching theoretical knowledge through voices
and echoes: a Vygotskian perspective’. In Proceedings of PME XXI, Lahti, Finland, pp.81-88.
Descartes, R. (1637): La Géométrie. [Translated from the French and Latin by Smith, E. & Latham,
M. (1954). Dover, New York.
Margolinas, C. (1993): De l’importance du vrai et du faux dans la classe de mathématiques, La
Pensée Sauvage, Grenoble.
Bartolini Bussi, M. (1995): ‘Voci della storia dell’algebra’, Actes du Seminaire Franco-Italien de
Didactique de l’Algèbre, I.M.A., Genova.
European Research in Mathematics Education I.II: Group 6 155

Bos, H. J. M. (1981): ‘On the representation of curves in Descartes’ Géométrie’, in Archive for
History of Exact Sciences , 295-338.
Capponi B, Laborde C. (1995): Cabri-classe, apprendre la géométrie avec un logiciel, Editions
Archimède, Argenteuil.
Capponi B. (1994): ‘Utilisation du logiciel Cabri-géomètre en classe’, Actes de l’Université d’été ,
“Apprentisage et enseignement de la géométrie avec ordinateur. Utilisation du logiciel
Cabri-géomètre en classe”, Grenoble.
Duval, R. (1988): ‘Graphiques et equations: l’articulation de deux registres’, Annales de Didactique et
de Sciences Cognitives, vol I. [Spanish translation: ‘Gráficas y Ecuaciones: la articulación de
dos registros’, by Parra, B., in Antología de Educación Matemática, CINVESTAV-IPN.
Fuller, G. (1979): Analytic Geometric, Addison-Wesley Pub.Company, Inc. [Spanish translation:
1981, Geometría Analítica , Compañía Editorial Continental, México].
Gillies, D. (Ed.) (1992): Revolutions in Mathematics, Clarendon Press, Oxford.
Herscovics, N. (1980): ‘Constructing meaning for linear equations: a problem of representation’,
Recherches en Didactique des Mathématiques, I (3), 351-385.
Hoyos, V. (1996): Del pensamiento algebraico operacional básico al pensamiento algebraico
analítico, Tesis de Doctorado, CINVESTAV-IPN, México.
Hoyos, V. (1997): Algunos de los problemas y teoremas vinculados con los orígenes de la geometría
analítica , Tesis de Matematicas, Fac. de Ciencias, UNAM.
Hoyos, V. (1998): ‘La signification des variables algébriques: une étude empirique auprès de lycéens
(16-18 ans)’, in Didactique et technologies cognitives en mathématiques -Séminaires 1998, (To
Kuchemann, D. (1981): ‘Algebra’, Chap. 8, in Children’s understanding of mathematics: 11-16.
(Coord. of K. Hart), John Murray, London.
Laborde, C. (1993): ‘The computer as part of the learning environment: The case of géométrie’, in C.
Keitel & K. Ruthven (Eds.), Learning from Computers: Mathematics Education and
Technology, Springer-Verlag, Berlin.
Laborde C., Capponi B. (1994): ‘Cabri-géomètre constituant d’un milieu pour l’apprentissage de la
notion de figure géométrique’, Recherche en didactique des mathématiques, 14, 1-2.
Laborde J-M., Bellemain F. (1998): Cabri-géomètre II . Windows Software of Dynamic Geometry .
CNRS-Université Joseph Fourier, Grenoble, France; Texas Instruments, Dallas, Texas USA.
Noss, R. & Hoyles,C. (1996): Windows on Mathematical Meanings, Kluwer, Dordrecht.
Schoenfeld, A., Smith III, J., and Arcavi, A. (1993): ‘LEARNING. The microgenetic analysis of one
student’s evolving understanding of a complex subject matter domain’, R. Glaser (Ed),
Advances in Instructional Psychology (Vol. 4). L. Erlbaum, Hillsdale, NJ.
European Research in Mathematics Education I.II: Group 6 156

1. In Hoyos' observation (1996) in the task of making the graphical drawing of the
straight line 2x + y = 16; the students start to reduce the given equation and obtain
correctly y = 16 - 2x. But they had not managed the calculations from the reduced
PB: To give values to x: look, if I give a value to x , two; it is going to be 4, plus ,
plus y, equals 16 minus 4, it means that y is 12.
Thus, one of the cases (PB) indicated the equation 2x+y=16 while he was
calculating out loud and after he had indicated y = 16-2x, finally he wrote y=12.
PB: But I need another here I put 4; multiplied by 2 is 8; 2 multiplied
by...8, plus y equals 16, 16...
When x is 4, y is 8. And here I already have 2 points to draw the straight line.
(PB case interview, annexe 5 in Hoyos, 1996).
We have to note that, this time again, the student PB was calculating by putting up
with the form 2x+y=16 and not from y=16-2x.
2. Windows' Version of Cabri II. (This version is provided by Texas instruments).
3. Lycée Aristide Bergès Seyssinet-Pariset 38170. France
4. By the way, we have made a study (Hoyos1996) about the emergence of algebraic
equations in mathematics' history. We think that Descartes with his Geometry
(Gillies, D. (Ed.), 1992), Revolutions in Mathematics, Clarenda Press, Oxford
science publications (USA)) has accomplished one of the mathematical
resolutions, the one that represents algebraically some curves named
geometrically. According to Descartes, at the beginning of the second book of "La
Geometrie", it is possible that the reason why ancient geometers could not go
further in the study of curves more complex than the conic ones was that the first
considered ones where "the spiral, the quadratic and similar which belong to
nothing else but mechanics" (Descartes 1637, p317).
In "La Geometrie", Descartes shows us another way of considering the curves; a
new way completely revolutionary for the time:
« Et il n'est besoin de rien supposer pour tracer toutes les lignes courbes, que je
prétends ici d'introduire, sinon que deux ou plusieurs lignes puissent être menées
l'une par l'autre, et que leurs intersections en marquent d'autres... prenant comme
European Research in Mathematics Education I.II: Group 6 157

on fait pour Géométrique ce qui est précis et éxact, et pour Méchanique ce qui ne
l'est pas; et considerant la Géométrie comme une science, qui enseigne
generalement à connaitre les mesures de tous les corps, on ne doit pas plutôt
exclure les lignes plus composées que les plus simples, pourvu qu'on les puisse
imaginer être décrites par un mouvement continue, ou par plusieurs qui
s'entresuivent et don’t les derniers soient entièrement reglés par ceux qui les
precedent. Car par ce moyen on peut toujours avoir une connaissance exacte de leur
measure. » (Descartes, 1637, p.316)
« Je pourrais mettre ici plusieurs autres moyens pour tracer et concevoir des lignes
courbes, qui seraient de plus en plus composées par degrées à l'infini; mais pour
comprendre ensemble toutes celles, qui sont dans la nature, et las distinguer par
ordre en certaines genres, je ne sache rien de meilleur que de dire que tous les
points, de celles qu'on peut nommer Géométriques, c'est à dire qui tombent tous
quelque mesure précise et exacte, ont nécessairement quelque rapport à tous les
points d'une ligne droit, qui peut être exprimé par quelque équation, en tous par une
même. » (Descartes, 1637, p.319)
5. Ministère de l'Education nationale, de l'Enseignement Superieur, de la Recherche
et de l'Insertion Professionnelle. (1995). Mathématiques -classe de seconde,
classes de premières et terminales , séries ES, L, S-, France: Centre National de
Documentation Pédagogique.
6. According to al.(1997), « un 'echo' (it's) a link with the voice made
explicite through a discourse »; &, a voice (it's ) « some expression (that) represent
in a dense and communicative way important leaps in the evolution of mathematics
and science ». In the theoretical framing worked by Bartolini, M.(1995) &
Boero, al.(1997) they suppose that the introduction in the classroom of the
voices of the mathematical and science hystory « might (by means of suitable
tasks) develop into a voices and echoes game..» (Boero et al., 1997, p.81).
7. EIAH (Environnements Informatiques pour l'Apprentissage Humain), is a working
team of Laboratoire Leibniz at IMAG & Université Joseph Fourier, in Grenoble,
8. Cabri-géomètre II gives coordinates of each point.
9. Briefly explain the importance we give to work on the CABRI-II microworld, we
quote the words of Noss, R.&Hoyles, C., (1997, pp.132): « the system carries with
European Research in Mathematics Education I.II: Group 6 158

it elements of what is to be appreciated, it arranges activity so as to web solutions
strategies. The structures which form that web of meanings are, of course,
determined by pedagogical means: they are not arbitrary. There is human
intervention, mathematical labour, dormant within the system, and the learner has
to breathe life into in. »
European Research in Mathematics Education I.II: Group 6 159

Nicolina A. Malara1, Rosa Iaderosa2
Department of Mathematics, University of Modena, via Campi 213/b, 41100 Modena, Italy
Middle School “L. da Vinci”, Cesano Boscone (Milano), Italy

Abstract: After a brief introduction and a few general considerations on syntactic
difficulties in the interweaving of arithmetic and algebra, we analyse the conflict
between additive and multiplicative notation in arithmetic-algebraic realm and present
the first results of some activities carried out according to our hypothesis of research
with the aim of overcoming such conflict and promoting the semantic control of
complex writings. We conclude with some reflections concerning the choice and
difficulties of the didactic activities in middle school on these topics.
Keywords: algebra teaching, learning syntactical, structural aspects.

1. Introduction

There are several studies devoted to the problems connected to the passage from
arithmetic to algebra, and specifically to the pupils’ difficulties; scholars have different
opinions on the actual possibility of overcoming such difficulties through appropriate
teaching. Herscovics (1989) states that not all the difficulties that the pupils have on
passing from arithmetic to algebra are due to the kind of teaching, since many of its
obstacles derive from the way this passage developed historically. Harper (1987) too,
believes that the pupils’ learning repeats the historical process and therefore meets
obstacles and difficulties witnessed by the history of the developing of the algebraic

Nevertheless, many studies show that the difficulties in the approach to algebra are
often caused by a teaching of arithmetic focussing on the results of calculus processes
European Research in Mathematics Education I.II: Group 6 160

rather than on its relational/structural aspects (see the surveys by Kieran 1989, 1990,
1992 or by Malara 1997a, 1997b). The teaching of elementary algebra stresses on
symbolic representations (expressions, equations, functions) created by generalizing
the processes enacted by the mathematization of various situations (Sfard 1991, 1994).
This shifting from the process to the object is not sufficiently underlined in the teaching
and the unspoken changing of perspective gives the pupils great confusion and often
makes them accept passively the rules and techniques of formal calculus without any
control of the meaning conveyed by these rules or of the properties on which these
techniques are based.

The work that we present is included in one of our research projects (see Malara
1994, Malara et al. 1998), aimed at constructing with and for the teachers an innovative
didactic itinerary in middle school (pupils aged 11 to 14) that on the basis of the
research results and according to Italian syllabuses could allow a less traumatic and
more aware passage from arithmetic to algebra. The project aims at overcoming the old
Italian teaching tradition which introduces algebra ex abrupto in third grade and deals
with it only from the syntactical point of view. It promotes learning algebra as a
language (either in its syntactical aspects or in translation and of
production/communication of thinking) through and for the study of problems, even
those internal to mathematics and demonstrative ones. (see Malara &Gherpelli 1994
and 1997). Through the project we try to study if and to what extent a constructive
teaching and in particular an early introduction of letters in parallel with a constant
work of reflection and control of the meanings of which they are bearers- may limit or
even overcome well known obstacles and difficulties.

The work focuses on some important questions related to syntactical, relational and
structural aspects of arithmetic and algebra. More precisely, it concerns:
• the relational network (to be revised) between arithmetic and algebra in
‘elementary’ teaching;
• some aspects of arithmetic that can cause particular difficulties either within the
learning of arithmetic itself or of algebra;
• the introduction of classroom activities concentrating on the additive and
multiplicative notation as well as multiplicative/exponential (and their
European Research in Mathematics Education I.II: Group 6 161

combinations), whose difficulties should be highlighted by meaningful

It is not easy to write these questions in order, because are tightly intermingled.
However, we shall face them starting from the cultural and didactic problems they
bring about, presenting the activities we created for their potential overcoming, or at
least for focussing on the difficulties detected. The activities are still being
experimented in various classes and so have not been included in the project yet.

2. A Few Considerations on Syntactic Difficulties in the
Learning of Algebra

Moving from arithmetic to algebra definitely creates didactic and learning problems:
indeed you may see the algebraic code as a generalization of arithmetical symbols; still,
as didactic research has highlighted, there are a few differences between symbolism in
arithmetical realm and symbolism in algebraic realm. One reason for this fracture is the
multiplicity of meanings or roles that the same symbol (having a univocal meaning in
arithmetic) obtains within the algebraic language. See parentheses for instance: in the
language of arithmetic they are used only to indicate the priority of an operation over
the others when this priority contrasts with the convention; in the algebraic language
the symbol ( ) can play the same role, but can also be used as a mere barrier between two
signs you may not write one beside the other. This new function of parentheses often
brings the pupils to the mistake of using parentheses in their first function only: this way
they get to mistakes such as: (-2+5)x(-4)=+3x-4 or (-2+5)x(-4)=+3-4, which changes
completely the meaning of the operation required.

On the other hand, you need even better control in expliciting or not expliciting
some operation signs as soon as you start omitting completely the sign of multiplication
between two literal symbols. In numerical realm the insertion of the dot instead of the
sign x in general is accepted, but the fall of this tiny operational sign gives vent to
ambiguities and confusion. Even in a rather short and simple writing like 2ab,
sometimes the pupils insert the additive notation interpreting it as 2a+b. And should
they face even more complex writings, they would have much bigger difficulties in
European Research in Mathematics Education I.II: Group 6 162

managing all signs at the same time, be they explicit or unspoken, referring to numbers
or operations. Sometimes syntactic errors are not due to the lack of understanding of the
various situations considered one by one. As it generally happens, the simultaneous
management of a multiplicity of situations determines errors that would never occur
individually (Drouhard & Sackur 1997).

Even the signs ‘+’ and ‘-’ give vent to errors of interpretation: in the algebraic
language they don’t stand for operations only (addition and subtraction); they can be
part of the number itself, the relative number; the sign ‘-’ can be used as a unary
operator to indicate the passage to the inverse. This last option causes further problems
if you want to lead the pupils to a good awareness in the use of symbols: since they are
instinctively induced to consider numbers as natural numbers, they hardly accept that a
number can be expressed by a couple of signs, namely a sign and a natural number, and
belong to an additive structure in which each element has its opposite. They also find it
hard to accept the idea of the inverse of a given number in the use of rational numbers.
Some protocols show that there are pupils who describe the following algebraic
formula: -(a2b+5b) as ‘the difference’ of the product of the square of a and b added to
the quintuple of b, which means they don’t notice that in the subtraction there are two
terms and here there are not.

A typical example of the various meanings of the signs is given by
æ 1 1ö
ç + - 3 + ÷ x ( -5) where a parenthesis is only a separator the other one has a double
è 2 5ø
role, a sign ‘-’ represents a unary operator, the other one is part of the number.

Moreover, as underlined by Demby (1997) and Reggiani (1996), it must be said that
weaker pupils who can control the meaning of signs and conventions hardly yield to
formal simplification and generally get back to the additive model, as shown in the
following examples (reported by Demby): i) the expression -2x2+8-8x-4x2 is
transformed into -2x2+8+(-8x)+(-4x2); ii) the expression (-4x+3)-(-1+2x) is
transformed into (-4x+3)+(-1)(-1+2x) or even into (-4x+3)+(1-2x).

Another thing that should be considered, as underlined by Booker (1987), is the fact
that, despite decomposition into factors, multiplicative representations of numbers are
rarely used to carry out operations in arithmetic, whereas this kind of representations
European Research in Mathematics Education I.II: Group 6 163

are often used in algebra as soon as letters are introduced. The operation of power on
which such representation is based, gives remarkable problems in arithmetical
activities and confirms its difficulties when operating on literal symbols. The
impression we have is that the presence of letters emphasizes these obstacles to the
extent of making it impossible to go any further, which is also due to the fact that the
absence of numbers does not allow to sidestep the application of the properties of
powers by resorting to ‘naive’ calculations of numerical kind. See, for instance, the
different strategies (both wrong) enacted by the pupils in the following two situations:
(22x32)3=(8x9)=72 ; (a2xb2)3 = a3xb6. Moreover, in the case in which a multiplicative
writing introduces a factor with unitary exponent, errors appear also in the numerical
case. The expression (2x32)3 is transformed into 2x36. It seems that the pupils multiply
the exponents only if there is one of them already explicit for the basis; the number
lacking in exponent is neglected and writes once again without expliciting the
exponent. Yet, in the numerical case, many overcome such kind of situations by
developing the product and the power within the parentheses and avoid to face the
power of power. In the literal case this is impossible and the mentioned error occurs in
grater percentage.

A further difference regards the sign of equality. Kieran (1990) underlines the
different meaning that it has in arithmetic and algebra: in arithmetical realm pupils
usually consider it to be a directional operator with the meaning of “gives vent to”
(which is proved by the reluctance with which they watch expressions like 4+3=6+1);
in algebra instead, even if this meaning appears in the simplification of expressions, the
meaning of relation of equivalence prevails and it achieves significance in the study of
equations (Herscovics & Linchevski 1992 and 1994).

This difference in dealing with literal and numerical expressions, with the
consequent errors, appears evident when you work on the two levels at the same time.
When the pupils, for example, were asked to operate simultaneously on the two
expressions (2x32)3+(22x5) and (ab2)3+(a2b) by using the properties of powers as much
as possible, 50% of the class avoided the obstacle as follows: in the arithmetical case
they calculated the product that is at the basis of the power first and obtained the right
result; whereas in the literal case only 10% of the pupils operated on powers by
European Research in Mathematics Education I.II: Group 6 164

elevating each factor of the product within the parentheses to the more external
exponent, all the others just got stuck in front of the exercise.

3. The Confusion Between Additive and Multiplicative
Notation: A Hypothesis of Inquiry for its Overcoming

We believe that for most of the syntactic errors detected it is possible to sort out a few
basic reasons that allow a unitary vision of the problems that the pupils have. Among
them, the main reason, which is also at the basis of our hypothesis of research, is the
following: the additive structure of the naturals and the consequent additive notation
constitute a very strong primitive mental model for the pupils, particularly in primary
school: the multiplicative model, which is introduced right afterwards, is not as strong
and often it doesn’t overlap correctly to the additive: it is less spontaneous, even
because it is learned later. Didactic research offers many examples about the prevalence
of the additive model in problem solving. Still, here we are talking about the prevalence
of the operation of addition as to multiplication not only semantically, but also from a
formal point of view. The pupils focus their attention on the sign ‘+’ and this leads them
to see ‘+’ in the situations in which the operative sign is not explicit (as we said,
sometimes the simple writing 2ab is interpreted as 2a+b). Moreover, as emphasized by
Fishbein (see Fishbein & Barach 1993 or Fishbein 1994), owing to their formal
elegance and visual incisiveness, some additive laws, such as m(a+b)=ma+mb,
overcome their status of laws and become strong models for the pupils, who extend
them inappropriately in multiplicative realm, confusing the roles of addend and factor.

The additive notation dominates and gets confused with the multiplicative one in
many situations. For example during a first grade activity in which the pupils were
asked to do transformations in order to calculate mentally more quickly, some pupils
transformed 148x20 into 148x10+148x10. This testifies that they could more easily see
20 as 10+10 rather than as 10x2. When asked to find the inverse of the rational number
3/5, some pupils declared they had been thinking of another fraction like 1-3/5, which
means they confused the inverse fraction with the complementary one. However, the
field in which these difficulties are more evident is where you operate with powers and
their properties. The product or quotient of powers with the same basis requires the
European Research in Mathematics Education I.II: Group 6 165

simultaneous management of both operations and notations: the multiplicative level
between powers turns into the additive one between the exponents, to which occurs the
difficulties due to the management of the power as a binary operation. Unlike the other
arithmetical operations that the pupils know ever since primary school, it is an
operation in which the sign is not explicit; moreover, the role of the two elements of the
couple is not the same: the exponent is just the index of how many times you have to
write the same factor (the basis) in a repeated multiplication. Even if the various
calculating techniques are learned by the pupils in a preliminary phase, the complexity
of an expression that simultaneously contains more operations of the same kind raises
lack of control over signs and meanings. The question can be expressed as follows: if
the conflict that the pupils experience between additive and multiplicative notation can
cause errors or confusion in the learning of the algebraic language and of its meanings,
is it didactically useful- and to what extent - to find out situations and strategies to
force the comparison and exchange between additive-multiplicative situations and
multiplicative-exponential situations?

The research we started with second- and third-grade pupils (teacher: R. Iaderosa)
moves right from this assumption and aims at analysing the pupils’ ability to spot out
(by intuition) the operative “structure” transferring it from the additive field to the
multiplicative one, and backwards.

Comparison between the multiplicative and additive realms
Middle school, second grade (pupils aged 12-13).
1) By using any time it is possible the properties of powers, calculate:
a) 33 x 32 + 33 + 32 ; b) 5 x 2 + 52 ; c) (2 x 32)2 + ( 2 + 32) x 2 .
2) Transform the following writings by replacing each sign of addition with a sign of
multiplication, and each sign of multiplication with a power:
a) (2 x 3) + 5 + 7 x 2 ® ; b) (2 x 5) x 7 + 2® ; c) (5 + 2) x 4® .
3) Transform the following writings by replacing each sign of multiplication with an addition
and each power with a multiplication:
a) 23 x 5 x 72 ® ; b) 53 x 24 x 3 ® ; c) (5 x 8) x 2 ® ; d) (52 x 22) x 3 ® .
4) Look at the following writings and say whether each of them is true or false:
a) 2 x 5 + 3 x 4 = 5 x 2 + 4 x 3 ; b) 25 x 34 = 52 x 43 .
Look at a) and b) once again. Is it possible to pass from one to the other like in the previous
European Research in Mathematics Education I.II: Group 6 166

Middle school, third grade (pupils aged 13-14).
By using as much as possible the properties of powers, calculate:
a) 23 x 22 + 23 + 22 a3 x a2 + a3 + a2
b) (2 x 3)2 + (23)2 (ab)2 + (a3)2
c) (2 x 32) + (22 x 5)2 (ab2) + (a2b)2

Tab. I

The activities, purposely arranged, have the aim of forcing the comparison between
the multiplicative and additive notations by opposing them in complex situations in
combination with the analogous multiplicative and exponential notations, with the aim
of improving the distinction between the two realms and spotting out analogies and
differences among their properties. The first productions by the pupils showed that 50%
of them could keep good control over the two notations. As foreseen, they had
difficulties in operating with powers and some of the weaker pupils simply tried to
explicit powers as products. In particular, while they understand pretty well the
correspondences (direct and inverse) between addition and multiplication, they cannot
control the operation of power at the same time, especially in the inverse passage, as
shown by the following examples of solution of the exercises in table 1: i) exercise 2a)
(2x3)+5+7x2 was transformed into: (23)5x72 ; 23+5x72 ; 23x5x72 ; 2x3x5x7x2; ii) exercise
3a): 23x5x72 was transformed into: 23+5+72 ; (2x3)x5x(7x2); (2+3)+5+(7x2). However,
after a few months of activity on these aspects, the pupils showed better ability in
grasping the differences between the two notations and a wider control of even more
complex situations. Some interesting protocols on these activities will be showed
during the presentation.

4. Considerations and Problematization of Some
Arithmetical-algebraic Activities in Middle School

As a matter of fact, algebra - seen as study of relations and procedures - begins and must
begin from the study of arithmetic. Still, to what extent the knowledge of arithmetic can
help when shifting to algebra? The relations between the two fields are much more
complex. For example, a deeper knowledge of arithmetic, of the formal properties of
European Research in Mathematics Education I.II: Group 6 167

the arithmetical operations, of the various numerical realms can favour the correct
learning of literal calculus and of some concepts tied to equations, algebraic operators,
etc. Nevertheless, it is the algebraic code itself that allows the comprehension of higher
level arithmetic. For example, the generalization of a rule or procedure that would
hardly be understood through a single listing of numerical cases can be expressed in a
literal code. The same way, some arithmetical properties are deduced from a literal
formalization. Paradoxically, in some cases the algebraic language can simplify the
analysis of arithmetical contents: it is well known that many of the properties
concerning the relation of divisibility between integers or the structure of some natural
numbers as to their factorization escape the numerical form and are highlighted by a
literal writing.

Particularly refined aspects of the teaching of arithmetic are implicit in problems
concerning the calculus with powers. In fact, if we consider the power at integer
exponent, defined in R, and the multiple after an integer of a real number, we can grasp
more easily the analogy between the two definitions and how in both situations we may
not speak of commutativity, since the two elements of the ordered couple are not seen
as belonging to the same domain. Unfortunately, at this level of schooling, when pupils
only just approached such topics and therefore do not operate on R but on N, it is
objectively very difficult for the teachers to make them distinguish the two levels. It is
therefore clear that the operation of power itself (usually introduced in first grades) for
many reasons can be hardly and definitely ‘unnatural’ to first-grade pupils who by no
means possess such algebraic tools of analysis. As to this, one could try to highlight the
analogy between the following two writings: 2x3 and 23. By formally representing the
two writings as: (2,3 ) ® 2+2+2 (where 2 is written 3 times); (2,3 ) ® 2x2x2 (where 2 is
written 3 times)

It is evident that the analogy consists in the fact that in both cases only the first
element of the couple is involved in the calculation, whereas the second expresses only
the number of times the addend or the factor is written.

So, it is not a matter of ‘doing’ algebra only after having introduced arithmetic, but
rather of reading and developing didactically the algebraic aspects of arithmetic, which
shouldn’t necessarily be delayed to the ‘end’ of the arithmetical contents. This should
European Research in Mathematics Education I.II: Group 6 168

be born in mind especially for the middle school, when teachers and pupils must not
pursue the technical skills that allow to deal new contents at superior level, and
therefore can find the occasions for reflecting systematically on these questions.
Indeed, often it is the ripening of the pupils’ thinking that allows them to overcome
some cognitive obstacles; still, there are concepts upon which they have never reflected
enough, even when they get to university.

The introduction of letters focuses on the aspects of arithmetic that the pupils hardly
ever interiorize: first of all the operation of power and the recognition of its properties,
the distributive property, the concept of inverse and opposite of a number, seen as
numbers first, and then as operators. And in fact it is from these situations that the
syntactical problems arise, since the limited and recent arithmetical experiences have
not allowed the appropriate interiorization. What has been learned is not an automatism
yet, whereas in this field one should, to quote Bell (1992), “make what has been
learned automatic and vice versa recognize in the automatism its meaning”. Let us
consider situations like the following: (2x3)2+2(2x3) and (ab)2+2ab. In the numerical
case, almost inevitably, in the pupils’ minds the development brings to the form 62+2.
In the literal form, operating with powers brings to the form a2b2+2ab, which explicitly
suggests the application of the distributive property and leads to the expression abx(ab +
2). The reading of the two formulas obtained is definitely different, and this should
make the pupils reflect ion the fact that the letters highlight relations and forms that the
numerical aspect somehow hides. This confirms the need of operating on the possible
equivalent expressions of a same number or of a numerical expression, as emphasized
in Malara (1994).

On the other hand, it is also necessary to consider the following points:
• how can a formal property for an arithmetical operation be fully understood if the
pupils still cannot abstract from single numerical example an equality among
procedures and express it in a literal form?
• how can they recognize an inverse element as to the multiplication in numerical
realm, if they still find it hard to consider fractions to be numbers?

On gradually introducing the use of letters instead of numbers, we saw that the same
paths and properties that seemed to be known, often are not recognized as such if they
European Research in Mathematics Education I.II: Group 6 169

appear with a new, or different, symbolism. This is evident as regards the recognition
and application of the formal properties of the arithmetical operations. Let us consider
the following example in which we report the various steps, in order of difficulty, that
should be explicited and understood in numerical realm in order to get to recognize and
apply correctly the distributive property in transforming 2ab+b into (2a+1)b: i)
3x2+5x2=(3+5)x2; ii) 3xa+5xa=(3+5)xa; iii) 2xa+ax3=(2+3)xa; iv) 2xa+axb=(2+b)xa; v)
2xa+a=(2+1)xa; vi) 2ab+b = (2a+1)xb. Protocols on the difficulties in facing these steps
will be showed in the presentation.

Thus, our belief is that only by carrying out a metacognitive teaching of arithmetic,
introducing early and gradually the use of letters and dealing from the very beginning
with the algebraic aspects which are contained in arithmetic itself is it possible to guide
the pupils towards the achievement of a correct use of the algebraic code, and in the
meantime start their acquisition of the relational and structural aspects which in future
perspective shall bring them to possess, at the end of high school, the tools to
comprehend the study of abstract algebra.

Finally, a special mention goes, as already underlined by Gallo (1994), to the
problem of control. Probably this problem has a different relevance and connotation for
middle-school pupils from older ones. Very often, the lack of control is not due to a
separation from the meanings, but more simply to the lack of autonomy in the
application of what has been learned. Many pupils find themselves in the ‘zone of
proximal development’ and need to be guided along the reproduction of an itinerary
that was learned in school, they simply cannot operate correctly in an autonomous way.
Moreover, when this control exists, if it works in an acceptable way in situations in
which the difficulties signalled by the literature are ‘diluted’ in a wider numerical or
literal realm, it turns out to be inadequate in situations in which you have to manage
more than one ‘crucial’ question at a time. But we believe that presenting these
questions systematically in complexity can emerge a real control, unrelated to the
acquisition of mere (though correct) automatisms: an almost automatic awareness.
European Research in Mathematics Education I.II: Group 6 170

5. References
Boero P. (1992): Sulla specificità delle ricerche in didattica della matematica: il caso del formalismo
algebrico, L’Insegnamento della Matematica e delle Scienze Integrate, 15, n. 10, 963-986.
Bell A.W. (1992): Algebra Learning Research and the curriculum, proc WALT , Torino.
Demby A. (1997): Algebraic procedures used by 13 to 15 year old, Educational Studies in
Mathematics, 33, 45-70.
Drouhard J.P., Sackur C. (1997): Triple approach: a theoretical frame to interpret students’ activity in
algebra, Proc. of PME XXI, vol. 2, 225-232.
Fishbein (1994): The interaction between the formal, the algorithmic, and the intuitive components in
a mathematical activity. In R. Biehler et al. (Eds.) Didactics of Mathematics as a Scientific
Discipline, 231-235, Dordrecht: Kluver.
Fishbein E., Barach A. (1993): Algorithmic models and their misuse in solving algebraic problems,
Proc. of PME XVII, vol.1, 162-172.
Gallo E. (1994): Algebraic manipulation as problem solving. In N.A. Malara and L. Rico (Eds.), First
Italian-Spanish Symposium on Mathematics Education, 131-137, Modena University Press.
Harper E. (1987): Ghost of Diophantus, Educational Studies in Mathematics, 18, 75-90.
Herscovics N. (1987): Cognitive Obstacles Encountered in the Learning of Algebra. In S. Wagner and
K. Kieran (Eds.), Research Issues in the Learning and Teaching of Algebra, LEA, Reston
Virginia, 60-86.
Herscovics N., Linchevski L.(1992): “Cancellation within-the-equation” as a solution procedure,
Proc. of PME XVI, vol. 1, 265-272.
Herscovics N., Linchevski L. (1994): A Cognitive Gap between Arithmetic and Algebra, Educational
Studies in Mathematics, 27, 59-78.
Kieran K. (1989): The Early Learning of Algebra: a Structural Perspective. In S. Wagner and K.
Kieran (Eds.), Research Issues in the Learning and Teaching of Algebra, LEA, Reston Virginia,
Kieran K. (1990): Cognitive Processes involved in Learning School Algebra. In P. Nesher and J.
Kilpatrick (Eds.), Mathematics and Cognition, ICMI Study Series, Cambridge University
Press, 96-112.
Kieran K. (1992): The Learning and Teaching of School Algebra. In D.A. Grouws (Ed.), Handbook of
Research on Mathematics Teaching and Learning, 390-419, New York: Macmillan.
Malara N.A. (1994): Algebraic Thinking: How does it promote starting from compulsory school and
limiting its difficulties?, L’Educazione Matematica, anno XVII, serie V, 1, 80-99.
European Research in Mathematics Education I.II: Group 6 171

Malara N.A. (1997a): Problemi nel passaggio Aritmetica-Algebra, La Matematica e la sua Didattica,
2, 176-186.
Malara N.A. (1997b): Rassegna delle ricerche sui problemi di insegnamento-apprendimento
dell’algebra in ambito PME (A survey of PME researches on the problems of teaching-learning
of Algebra) report presented at the “IX Seminario Nazionale di Ricerca in Didattica della
Matematica”, Pisa December 1997.
Malara N.A., Gherpelli L. (1994): Argomentazione in Aritmetica (Argumentation in Arithmetics), in
Basso M. et. al. (Eds.) Numeri e Proprietà (Numbers and Properties), 55-60. Parma Univ. Press.
Malara N.A., Gherpelli L. (1997): Argumentation and Proof in Arithmetics: Some Results of a Long
Lasting Research, in L’Educazione Matematica, anno XVIII, serie V, 2, n. 2, 82-102.
Reggiani M. (1997): Continuità nella costruzione del pensiero algebrico (Continuity in the
Construction of Algebraic Thinking), Notiziario UMI suppl.n.7, 35-62.
Sfard A. (1991): On the Dual Nature of Mathematical Conceptions: Reflections on Processes and
Objects as Different Sides of the Same Coin, Educational Studies in Mathematics, 22, 1-36.
Sfard A. (1994): The Gains and Pitfalls of reification: The case of Algebra, Educational Studies in
Mathematics, 26, 191-228.
European Research in Mathematics Education I.II: Group 6 172

Maria Reggiani
Department of Mathematics, University of Pavia, via Abbiategrasso 215, I-27100 Pavia, Italy

Abstract: The capacity to write and to solve equations is a crucial point in the
construction of algebraic thinking, since it involves both the mastery of the formal rules
of algebraic language and calculation, as well as the correct interpretation of the
meaning of the symbols used. Our research study is mainly concentrated on examining
the possibility of improving the capacity of students to solve first grade equations
during their first year of high school, by establishing a dialectical relationship between
the application of properties and their semantic control. In this report we would like to
present certain observations concerned with the problem exposed in relation to
equations without solutions or with an infinite number of solutions.
Keywords: symbols, rules, meanings.

1. Introduction

Research into teaching and learning algebra has demonstrated that one of the
fundamental problems is the students’ difficulty in being able to manage a formula and
its meaning at the same time. (Arzarello et alii 1995, Malara 1997, Sfard 1991, Cortes
1993, Linchevski & Herscovics 1994,…)

When considering in particular the capacity to write and solve equations, numerous
research studies have brought to light various other aspects which are included in this
process. In particular, studies have been carried out concerning certain problems linked
with using the equal sign in its relational sense (Sfard 1994) and the various uses of
letters as unknown quantities and as parameters (Ursini 1997). One of the fundamental
aspects which cannot be ignored when working with equations, is the fact that while the
equation in general represents the symbolic translation of a problematic situation, and
European Research in Mathematics Education I.II: Group 6 173

the solution of the equation leads to the solution of the problem, the various
intermediate passages cannot always be easily read as the stenographic transcription of
the resolving strategy (Boero 1992). It should also be remembered that numerous
studies have demonstrated the fact that students of 11 to 12 years of age, and sometimes
even primary school pupils, are already capable of successfully applying linguistic type
or ”elementary” method strategies (inverse operators , graphs, etc...) to solve equations
which are linked (or not) to a problem situation (Reggiani 1994, Cortes 1993).

Our research involves the study environment which in classical terms could be
defined as the syntax - semantic relationship, with special interest dedicated to the
problem of semantic control over the operational methods (Gallo 1994). This work also
takes into consideration the result of research carried out previously by our group on
problems connected with the passage from arithmetic over to algebra - and especially
with reference to the problems of the generalising and the use of symbols and
conventions in algebraic language. Our observations have shown that alternating the
use of different environments (arithmetic, computer information, geometry) for an
approach to algebra seems to be an efficient method for making semantic control of
algebraic operations easier.

2. Research Context

From the results observed after certain secondary school entrance tests, we focussed
our attention on some of the answers given by pupils to questions which demanded the
solution to simple equations, directly or in the form of inverted formula. An analysis of
the results has shown performance levels much lower than work carried out by the same
students in other parts of the same test. According to literature, we observed that in
these cases, the aspect which was lacking was the control of the results and of the
transformations carried out.

So it was decided to study the possibility of improving the pupils’ capacity during
their first year at secondary school (14 year olds) as far as solving equations was
concerned, improving the semantic control both with respect to the context in which the
problem is posed (for example in the case of inversion of formulas, or of equations used
European Research in Mathematics Education I.II: Group 6 174

to solve a problem) as well as with respect to the meaning of the algebraic operations
made by the pupils themselves.

In fact, we feel that the solution to an equation, which requires good competence in
applying the formal rules, represents the synthesis of the passage from arithmetic to
algebra only if a process of generalisation (abstraction) has been achieved which refers
to the meaning of the objects of the algebraic language, and if the pupils have
understood the semantic implications of the formal rules and the operative conventions
of algebraic language.

Often, as shown also by some of our previous research studies, these capacities have
not been acquired and this results from the fact that the pupils try to replace them with
”rules” learnt in a mechanical way. Through a large number of examples introduced at
teaching level (with the use of diagrams, verbal explanations, class discussions, etc.)
our research concentrates on identifying the point of equilibrium between operating
methods and the meaning of the operations that the pupils are working on - that is:
between syntactical and semantic aspects - while being well aware of the difficulty of
improving the learning capacity on both levels: in fact, as several authors have pointed
out, very often the excessive attention to meaning slows down the necessary training in
self control needed for the pure application of the rules for manipulating expressions

3. Research Problem

In the environment of the problems described in the previous paragraph we are studying
the persistence of the fracture between operations and meaning in algebraic contexts
where formalism (linked to the capacity for abstraction) becomes more important, in
particular when dealing with equations without solutions, with infinite solutions, or
with ”unacceptable” solutions.
European Research in Mathematics Education I.II: Group 6 175

4. Research Method

The work phase described in this report was carried out in the following manner: in a
first year secondary school class composed of 27 pupils, during normal scholastic
hours, a teaching program of 15 hours was set up between the teacher and the research
group on the subject of first grade equations and based on the observations which had
emerged during previous research by our group (Bovio et al. 1995). Also present during
the lesson was a student in her last year of a Mathematics degree as an observer.

At the end of the program a set of written and individual tests was carried out in
order to check the level of competence the pupils had acquired in relation to first grade
equations, and in particular, in situations where it was necessary to keep in mind the
meaning of the operations being worked on. The analyses provided in this report refer
to the individual protocols of the pupils.

5. Summary of the Work Carried out in Class

Since the discussion concerning the protocols under examination cannot be separated
from the activities carried out beforehand, a short summary of the whole learning
program is essential.

The work in class was concentrated on the following points:
• Recovery of the pupils’ ability to solve problems through equation using
elementary methods in order to establish a connection between the work
proposed and the activities carried out by the pupils previously (generally during
intermediate school). Through observation of spontaneous strategies this stage
was directed at attributing meaning to equation solving procedures. In fact, these
procedures often compose the generalisation for elementary methods of solution
(the use of inverse operators for example, or using the example of balancing
scales, or the simple ”verbal solution” often used in the simplest cases), but often
the pupils do not establish any connection between procedures and spontaneous
European Research in Mathematics Education I.II: Group 6 176

• Solutions, using elementary methods or through attempts, of equations even on
levels higher than first grade equations which are easily factored or are not
polynomial (for example exponential), with the aim of generalising the meaning
of the term equation and to clarify the meaning of the term solution.
• Solution to first grade equations through operations of transformation by trying
to underline the meaning of the transformation by emphasising the necessity that
transformations are inverted. During this stage, we also used tables in order to
provide a kind of graphical diagram to the solution of the equations through
transformations, and at each passage, emphasising the set of the solutions and
comparing that set with the one identified previously. Obviously during this
stage we spent some time on the problem of the ”zero”, comparing its different
roles when it is an additive or a multiplicative term or a denominator. We
discussed with the pupils the fact that equations can have a finite number of
solutions, (underlining in particular the cases where there is only one solution) or
can have no solution, or an infinite number of solutions.
• Discussions of the problem of the presence of letters which are different from the
unknown in an equation, and how to deal with denominators.

In agreement with the idea of learning proposed by the theories based on
constructivism the work methods in class were based on the individual study of
problems and then on the following group discussion on the solutions proposed by the

6. Verification Tests

The tests run at the end of the program were composed of various verification checks
which included two problems expressed verbally to be converted into an equation and
(subsequently) solved; the solving of certain first grade equations of various levels of
complexity, one with letter coefficients; and certain equations not belonging to the first
grade, to be solved using elementary methods.

Here we would like to focus a particular problem:
European Research in Mathematics Education I.II: Group 6 177

”Consider the equations of each of the following groups and reply whether they have
the same solutions, giving reasons for your answers:
a) 2x-7 = 5 2x = 5+7 2x = 5-7 x-7/2 = 5 x-7/2 = 5/2
b) 0·x = 5 x=5 x=0 x = 0/5 0·x = 10
c) x = -3 0·x = -3·0 x+2x = -3+2x x·x = -3x
d) (x+2)/3 = (x-5)/2 2x+2 = x-5 x = x-7 0·x = -7 “

7. Analysis of the Proposed Problem

The four groups of equations represent different situations. In each group there are
equivalent equations and non-equivalent equations. The equations are generally
obtained from each other by applying the usual transformations (adding or subtracting,
multiplying or dividing both members by the same quantity); however in some cases
we reproduced typical mistakes, obtaining non equivalent equations.

As we mentioned before, the pupils from the previous teaching program have
worked on these transformations identifying the limits of application; repeated use was
also aimed at establishing them as rules. It is necessary that the solution of a first grade
equation becomes (at least in part) one of the automatisms necessary to be able to deal
with more complex problems.

The object of the exercise therefore is to verify the pupils’ capacity to identify the
cases where it is not natural (or is incorrect) to use the rules they have learned. Based on
the kind of didactic activity previously carried out, we feel that when pupils make this
choice they should be guided by the meaning of the operation they are working on. In
particular, in the class discussions, we insisted heavily on the meaning of ”resolving an
equation in a set” - understood as : ”How to find the values which replace the unknown
in order to make the equality true”; during the discussion we also underlined that only
the transformations of an equation which maintain that it is true for the same values,
without adding others, may be accepted.

Of the four groups of equations proposed:
European Research in Mathematics Education I.II: Group 6 178

The equations of group a) checked only the capacity to apply correctly
transformations normally used to solve a first grade equation with numerical
coefficients where the unknown appears only once.

The equations in b) checked the capacity to recognise first grade equations without
solutions, to see that in this case the fact that the equations have ”the same solution”
does not depend on the fact that they are obtainable from each other using the normal
transforming methods. Alongside impossible equations, we included simple equations
which correspond with possible incorrect attempts to apply the usual rules. This group
of equations focalized attention on the role played by ”zero” with a certain amount of

The equations in c) reproposed the problem of multiplying by zero, checking the
pupils’ capacity to carry out cancelling operations in relation to terms including the
unknown .

The equations in d) lastly, proposed the problem of eliminating the denominator in a
simple case and faced the problem of equations without solutions in the case where ”x is

Each group of equations proposed different situations, and therefore it was not used
to check whether the pupils overcame a single difficulty. Since the check was at the end
of a program where we worked separately on the single aspects examined in this test,
we felt it was necessary to propose them in a more complex way, creating a new
situation, where the pupil was confronted with a context that did not reproduce the
teaching examples he had experienced previously in class, - in other words - exactly
what happens when one has to apply acquired knowledge in a different context.

However, these tests were compiled in such a way to permit pupils and teachers to
examine the various aspects singly.
European Research in Mathematics Education I.II: Group 6 179

8. Analysis of Results

Out of the 27 pupils who sat the test, only 8 solved the problem correctly and
completely, providing adequate reasons for their answers; 14 pupils obtained
intermediate results, indicating that on the whole, they were able to carry out the
transformations which conserve the solutions of an equation; on the other hand the
remaining 5 pupils showed serious problems.

In the next paragraph we will analyse the performance of the five pupils with
reference to their protocols.

As far as the other students are concerned, without providing a complete picture of
the results for each single item, we have limited the analysis by underlining the fact that
all the pupils answered point (a) correctly; almost all compare correctly the first
equation with the third and the fourth in point (c), and the first two in point (d).

Pupils had most difficulty with the equations in point (b) and also in the other
exercises where they had to establish the equivalence between equations without
solutions, or with infinite solutions.

Among the group classed as ”intermediate”, we noted that certain pupils do not
carry out a complete analysis of those points, or else - they give correct answers but
without any explanation; or else they make a mistake when resolving one comparison,
but carry out others which are just as difficult, correctly. Therefore we feel that in spite
of the fact that they tend to apply the ”rules”, even when solving problems where
attention must be paid to the ”meaning”, these pupils manage to keep situations even
more difficult than usual from a semantic point of view, under control.

Even keeping in mind the fact that certain aspects are incomplete, or lack
explanations, we can conclude that these pupils have perhaps acquired ”more complete
European Research in Mathematics Education I.II: Group 6 180

9. Analysis of Single Protocols

In this paragraph, we will analyse the protocols of the pupils who showed the greatest
difficulty. We will limit this analysis to those points described in the previous
paragraph as most important.

Pupil 1
(point b): ”Taking 0·x=5 I multiply both members by 0 and obtain 0·x=5·0 - that is x=0
which is equivalent to 0·x=5. But 0·x=5 is not equivalent to x=5 because the results are
different. x=0/5 is not equivalent to 0·x=5 because 0·x=5 is an identity while 0·x=10 is
equivalent to 0·x=5 because when both are multiplied by zero, both are equivalent to
For this pupil, all the equations of point (c) are equivalent because they are obtained
from one another by multiplying (or adding) both members by the same quantity. In
relation to point (d) the same pupil states that ”in the equation x=x-7, if you subtract x
from both members you obtain 0=-7 and not 0·x=-7”.
In the protocol of this pupil we notice the presence of mistakes in syntax such as stating
that 0·x=5·0 is equal to x=0, that is, that 0·x is not 0 but x. As is easily noted, this
conviction is persistent. It is also obvious that the pupil tends to apply rules and tries to
use terms which she has not yet understood the meaning (0·x=5 is an identity). We also
note the use (positive in itself) of different explanations for the various situations but
without operating any control over the coherence of the results and the explanations.
(0·x=5 is equivalent to x=0 because I can obtain both equations from one another; it is
not equivalent to x=5 because the results are different: for this pupil, what is the result
of 0·x=5, which as we noted, was first defined as an identity?)

Pupil 2
(point b) ”0·x=5 is equivalent to x=5/0 which is also equivalent to x=0. But x=0/5
cannot be equivalent to x=0 and therefore to the equation assigned initially”
(point c) “From 0·x=-3·0 I obtain x= -3·0/0, that is x=-3”
In the same way the pupil explains that ”from x·x= -3x I obtain x=-3”
As for point (d) the pupil limits her observations to the non equivalence of the first two
equations ”because the denominator is not the same”.
The pupil deals with all the equations by following the layout where the syntax solution
European Research in Mathematics Education I.II: Group 6 181

of the equation ax=b is x=b/a . In the first exercise there is also a syntax error which we
found frequently in the entrance tests: that is a/0=0 . The pupil also states that on the
contrary 0/a is not 0. In the second exercise the pupil simplifies 0/0=1.

Pupil 3
(point b): ”x=0 is equivalent to 0·x=5” (without explanations). No other answers are
given for this point.
All the equations of point c) are equivalent”
”The three first equations of point d) are equivalent”.

Pupil 4
(point b):”x=0 and x=0/5 are not equivalent to 0·x=5 because only the second member
is zero” The pupil made no comments on 0·x=10. In the same way for point (c) the pupil
establishes the equivalence between the other equations but does not take the second
into consideration without supplying any explanation.
In the same way in point (d) the pupil does not comment on the last example.
Pupils 3 and 4 probably suffer from the same difficulties as those already observed,
however the fact that they supply no explanations shows a behaviour pattern which is
perhaps even more confused when dealing with the cases where it would be necessary
to verify if it is possible to use the methods they have learnt.

Pupil 5
(point c): ”By multiplying by zero a false equality can be made true, so, by moving from
the first to the second equation, the equality has become true”
Concerning other points - the pupil repeated the same mistakes already met previously.
This pupil supplies an interesting explanation: ”The multiplication by zero makes a
false equality true”. However all forms of conclusion are missing as well as any type of
more precise connection with the case in hand. It would seem that the pupil repeats even
at explanatory level a statement which probably came up in class discussion during a
lesson but that he did not understand the exact meaning and limited himself to learning
it as another ”rule”.
European Research in Mathematics Education I.II: Group 6 182

10. Final Remarks

The presented test mainly gave positive results since it demonstrated that for most of
the pupils the learning process was activated: a teaching program, like one we have
described, concentrated on an ”operation-meaning” relationship, seems to have led the
pupils to applying the rules using a ”reasoning process” which permits most of them to
use the rules correctly in most cases.

However there are certain cases where a lack of success was very evident: the pupils
who are not capable of the effective learning process make an effort to apply rules but
are not able to exercise any control over the passages they do; this becomes clear
especially in the presence of zero, when it is essential that the pupil has a good
understanding of the meaning.

These pupils are also hindered by their difficulties with arithmetic - such as
confusion between multiplying and dividing by zero, or the fact that they are not aware
that 0a=0. Very often they try to repeat definitions or explanations at the very moment
when they are not able to attribute a precise semantic value.

This type of lack seems to be qualitative, in that it does not concern having or not
having worked out a certain number of operations to consolidate the use of certain
rules, but rather the inability to include them into a synthetic structure where each
element obtains its own value through its interrelation with the others.

This was already the aim of the teaching program carried out, but evidently these
pupils were not able to participate.

This does not mean that these pupils are excluded from the possibility of reaching
the goals which have been envisaged, but we think that results such as these will be
constructed (in their personal mental experience) not so much through the
sedimentation of operational practice, but more probably, suddenly in the context of an
adequately organised individual discussion based on exercises already carried out.

We feel that this type of evolution can be the object of a continuation of this research
European Research in Mathematics Education I.II: Group 6 183

11. References
Arzarello F., Bazzini L., Chiappini G.P. (1995): The construction of algebraic knowledge: towards a
socio-cultural theory and practice. Proceedings PME 19, L. Meira, D. Carraher (Eds.), Univ.
Federal de Pernambuco, 1, 119-134.
Bazzini L., Gallo E., Lemut E. (1996): Studies on Teaching and Learning Algebra, Italian Research in
Mathematics Education 1988-1995, Malara, Menghini, Reggiani (Eds.) Seminario Nazionale di
Ricerca in Didattica della Matematica, Consiglio Nazionale delle Ricerche, 40-55.
Boero P. (1994): About the role of algebraic language in Mathematics and related difficulties,
Rendiconti del Seminario Matematico, Torino, 52, n.2, 161-194 .
Bovio M., Reggiani M., Vercesi N. (1995): Problemi didattici relativi alle equazioni di primo grado
nel biennio delle superiori. L’Insegnamento della Matematica e delle Scienze Integrate, 18b,
n.1, 8-32.
Chiappini G., Lemut E. (1991): Construction and interpretation of algebraic models, Proceedings
PME 15, Assisi,.I, 199-206.
Cortes A. (1993): Analysis of errors and a cognitive model in the solving of equations, Proceedings
PME 17, I, 146-153.
Gallo E. (1994): Algebraic manipulation as problem solving, First Italian Spanish Research
Symposium in Mathematics Education, Malara, Rico (Eds.), Modena, 131-138.
Linchevski L., Herscovics N. (1994): Cognitive obstacles in pre-algebra, Proceedings PME 18, III,
Malara N.A. (1997): Problemi di insegnamento-apprendimento nel passaggio dall’aritmetica
all’algebra, La Matematica e la sua Didattica, 2, 176-186.
Reggiani M., (1994): Generalization as a basis for algebraic thinking. Observations with 11-12 year
old pupils Proceedings PME 18, IV , 97-104.
Reggiani M. (1995): Linguaggio algebrico e convenzioni: un’analisi del punto di vista degli alunni.
Séminaire Franco-Italien de Didactique de l’Algèbre, Documents des Séminaires SFIDA 1 à
SFIDA 4 1993 - 1995, Drouhard, Maurel Eds., IUFM Nice, IV 7- IV 14.
Reggiani M, (1997): Continuità nella costruzione del pensiero algebrico, Atti XVIII convegno
nazionale UMI-CIIM “Dalla scuola media alle superiori: continuità nell’insegnamento della
matematica” , 35-48.
Sfard A., (1991): On the dual nature of mathematical conceptions: reflections on processes and
objects as different sides of the same coin. Educational Studies in Mathematics, vol.22, 1-36.
Sfard A. (1994):The benefits and the pitfalls of reification: the case of equations and inequalities,
Rendiconti del Seminario Matematico, Torino, 52, n.2.
Ursini S., Trigueros M. (1997): Understanding of different uses of variable: a study with starting
college students, Proceedings of PME 21, vol. 4, 254-261.
European Research in Mathematics Education I.II: Group 6 184

Sam Rososhek
Department of Algebra, Tomsk State University, 36, Lenin avenue, Tomsk, 634050, Russia

Abstract: Research in how to form Algebra understanding has been done in frames of
MPI-project for many years. It is based on H.Weyl-I.Shafarevich conception about
Algebra as the collection of coordinatizing quantities systems. There are three main
informative lines in MPI-project systematic course of Algebra-functional, algebraic
structures and mathematical modelling. Two last of them are discussed here.
Keywords: algebraic structures, modelling, computer visualization.

1. Introduction

We present a part of the project of school mathematical education ”Mathematics.
Psychology Intelligence.”(MPI-project, Tomsk, Russia), the head of the project is Prof.
E. Gelfman. The project is directed at the development of students’ individual cognitive

The problem of Algebra understanding implies two other problems: the nature of
algebra and the nature of developing intelligence of a child. In short, teaching Algebra
should be organized in such a way as to reflect the role of Algebra as the human
intellectual culture phenomenon and, on the other hand, to provide the development of
individual intellectual abilities of a child.

2. Weyl-Shafarevich Conception

What is Algebra? Out of numerous possible approaches to answer this question, I
would like to attract your attention to H.Weyl - I.Shafarevich conception. H.Weyl : ”...
now we are coming back to old Greek viewpoint, according to which every sphere of
European Research in Mathematics Education I.II: Group 6 185

things requires its own numeric system defined on its own basis. And this happens not
only in geometry but in new quantum physics: physical quantities, belonging to a
certain given physical structure, permit themselves (but not those numeric values which
they may assume due to its different states), in accordance with quantum physics,
perform addition and noncommutative multiplication, forming by this some world of
algebraic quantities, corresponding to this structure, the world, which cannot be
regarded as fragment of the system of real numbers” [1].

I. Shafarevich summarized these ideas of H.Weyl in such a way :

a) every phenomenon, every process of real world (also in mathematics itself) may be
”coordinatized” in the frames of some system of coordinatizing quantities, that is
represented in some generalized system of ”coordinate axes”;
b) subject of Algebra is a study of various systems of coordinatizing quantities as
concrete (for example, numbers, polynomials, permutations, residue classes,
matrices, functions and so on)as well as abstract (groups, rings, fields, vector spaces
and so on);
c) if some phenomenon is not yet coordinatized by any familiar system of
coordinatized quantities, the problem of coordinatization arises. The task of algebra
is to solve this problem, that is to make up a new system of coordinatizing quantities.
Its further study may or may not be connected with problem of coordinatization,
with solving of which the given system of coordinatizing quantities appeared. [2]

3. Basic Elements of Algebraic Knowledge

This viewpoint turns Algebra from forum to explore the power of abstraction and
formal logic to the collection of coordinatizing quantities systems which are
meaningful for real world knowledge. Therefore we have to consider it necessary to
include the elements of abstract algebra in school curriculum. The next question arises:
how to form the understanding of Algebra by children? What requirements should a
school text meet for it? As for the informative aspect the following basic elements of
algebraic knowledge should be included in the school text organization:
European Research in Mathematics Education I.II: Group 6 186

1. Algebraic symbolics as the universal abstract language for describing reality.
2. Algebraic operation in the context of all its basic properties.
3. Algebraic structures as the specific form of coding information.
4. Algebraic notions semantics as the premise for the reality specific aspects
realization, aspects connected not only with the sphere of usual human experience,
but also with the sphere of human ”impossible” experience.

Thus, mastering algebra means mastering new language, new methods of
knowledge, new forms of the information organization, new reality view.

However for algebra to be perceived by a child in these most important aspects, the
psychological preparation of a child is necessary. It includes:

a) forming up conceptual experience gradually working with the subjective space of a
child’s understanding of separate signs meaning as well signs expressions;
b) developing maximally his imagination experience for the generalization as the
visualization ability allows to ”catch” the sense of mathematical notions without
any verbal-logical argument;
c) complicating his object operating experience, including both the main mental
operations and heuristic experiments and the mental experiment prediction;
d) forming the metacognitive experience of a child;
e) having the individual intuitive experience as the basis;
f) forming the multidimensional mental space for school material understanding.

4. Algebra Propaedeutics

Traditionally in Russia the systematic algebra course begins for children of 13. The
practical school experience shows that in spite of traditional previous preparation, there
exist some serious difficulties in the process of children’s understanding algebra. For
example, a psychological barrier of ”unwillingness” to accept algebraic symbolics
appears and children are not able to use algebraic language as the means of increasing
effectiveness of their intellectual work. Moreover, children begin to be afraid of
European Research in Mathematics Education I.II: Group 6 187

abstractedness of algebraic school material. We consider that many a difficulty is
explained by the fact that the informative and psychological lines do not equally exist in
algebra teaching. In consequence the cognitive experience of a child is not
reconstructed enough for a child to be psychologically prepared for the specific
character of algebraic material.

Gradual and prolonged introduction of letter signs in themes devoted to number sets
taught to 10-12 year old children in MPI-project1 has a definite aim: children must
intelligently estimate the role of a separate letter and algebraic expressions as an
intellectual instrument, making them able to fix and express relations between any
number objects in a visible and compact way, to discover the general way of solving
analogous tasks, to investigate hidden regularity in the sphere of this own subject and
number experience.

Some ways of methodical work aimed at the introduction of a letter sign and letter
expression in 5-6 grades (10-12 year old) lead to understanding that the central moment
which influences the progress of a skill development on the propaedeutics step is the
level of the formation of the number sign subjective image. The introduction of
algebraic symbolics on the propaedeutics step of algebra course certainly can’t be
reduced to the mere understanding of an algebraic expression as a specific language for
reality description by a child. It’s necessary for a child to get a skill of using different
problems. Thus a certain methodical problem emerges - to form a special style of
thinking of pupils. We make them know the main elements of problem solving with the
help of equations. They are taught to choose the variable, the basis for making up a
equation, the information necessary for problem solving under conditions of
expressiveness or insufficiency of data specially. We teach pupils to correlate the result
of equation solving with the word problem conditions, according to this equation, to
create different figurative notions of initial problem (in the form of drawing, schemes,
tables etc).

The pupils are taught to understand that one and the same real situation can be
expressed mathematically with the help of different equations and also that sometimes
outwardly different problems can be solved with the help of one and the same equation.
European Research in Mathematics Education I.II: Group 6 188

What concerns the psychological aspect of activity of a child, the preparation to
study algebra presupposes to form the skill to see regularity, to guess, to ask question, to
make hypotheses, to prove them and to contradict them, to work in the frames of the
pattern: how if it were...

5. Mathematical Modelling Line

The work using the algebraic symbolics as a special language for reality description and
for forming children’s special thinking style presents an opportunity to give various
problems to the pupils requiring attaching different mathematical apparatus while
organizing systematic algebra course. Here is, for example, one of the ecological
problems proposed for pupils of the 9 th grade (15 year old) in the textbook ” Equation

Foxes and rabbits live on the island. There is enough grass for rabbits to eat and foxes
eat only rabbits. It is required :
1) to find out how the amount of foxes and rabbits is to change at the initial moment of
time so that at the given moment their amount is the preassigned number;
2) to predict how many foxes and rabbits will be after the definite time.

Pupils come to know the scheme of mathematical modelling, developing the scheme
of its work while solving word problems. They get a skill of work with the obtained
model, to make charts of relation of foxes and rabbits amount to the time and also a
curve of interdependence of foxes and rabbits on the plane ”foxes - rabbits”. Then
pupils use the modelling skills answering the question: ‘’What if...?” For example they
find out the consequences of the island administration decision concerning the changes
in foxes and rabbits number at the initial moment, learn which of these decisions lead to
the destruction of the ecological system ”foxes-rabbits”. The increase of factors
number from two up to three, four and so on lead pupils to the necessity to solve linear
equation systems of corresponding order. Besides the mathematical modelling in this
case must be confirmed by the computer modelling. Systematic work with pupils in this
direction using ecological as well as economic models (salary - inflation, investments -
consuming, demand - price - supply etc) helps to form up multidimensional dynamic
European Research in Mathematics Education I.II: Group 6 189

thinking, which is so necessary for every school graduate in this complex quickly
changing world.

6. Algebraic Structures Introduction

The systematic course of algebra in MPI-project includes three main informative lines -
functional, algebraic structures and mathematical modelling2. Functional line is
traditional and we don’t dwell on it. Some description of mathematical modelling line
has been done before. Now we consider algebraic structures line.

The extension of number systems takes place in the 5 - 6 grades. Pupils come to
know fractions, decimal fractions, negative numbers. Every time they face the
necessity to find out the nature of new numbers, to study operating with numbers. We
think that at this step pupils can get the first experience of working with algebraic
operation as an object: they can learn the procedure of operations introduction and their
properties. One of the motives for new numbers introduction can serve the
impracticality of a definite operation on the ”old” number set. For example, the
introduction of negative numbers occurs while studying the practicality of the
subtraction operation on the natural numbers sets. Pupils should pay attention to the
fact if the new introduced operations have the properties of commutativity,
associativity, distributivity. Certainly it makes no sense to speak about any strict proof,
but we have an opportunity to create the experience of checking the presence or absence
of these properties.

The work with the number sets described above gives an opportunity later to
develop the functional aspect as well as the algebraic structures line in the systematic
algebra course. Realizing the functional aspect we used the pupils experience with the
operations on numbers, involving them into the search of the definitions of the
operations on monomials, polynomials, algebraic fractions. Developing the aspect of
algebraic structures we can go on analysing operations and objects worked at. So, we
want pupils to meet absolutely new for them operations. Some examples of such
operations have been done in the first book of the algebra systematic course ”Making
Friends with Algebra” (13 year old).
European Research in Mathematics Education I.II: Group 6 190

Task 1. Are whether the addition and multiplication the algebraic operations on the set:
a) odd numbers; b) even numbers?
Task 2. Is whether the subtraction the algebraic operation on the set:
a) naturals; b) integers?
Task 3. Is whether the division the algebraic operation on the set:
a) nonzero naturals; b) nonzero rationals?
Task 4. Check that the operation
is algebraic one on integers.
Is whether this operation associative and/or commutative?
Task 5. Check that the operation
x ( y = x + y – xy
is algebraic one on integers.
Is whether this operation associative and/or commutative?
Task 6. Make up some own examples of operations over numbers. Which from them are
algebraic one? Which from your algebraic operations are associative and/or

Such work with pupils allows children to get the experience of independent
constructing of algebraic operations with definite properties. Thus psychologically full
understanding of operations presupposes the children to use freely their knowledge
about operations even under unusual ”non-task” conditions. Such creative attitude to
an algebraic operation can be stimulated showing children the fact, that operations can
be fulfilled not only on numeric but on non-numeric objects as well.

The fact is that earlier children used the number notion in such problems which
required measurements. Moreover, every widening of a number notion was caused by
the necessity to measure one or the other value, which couldn’t be measured according
to the previous number notion. The same line is continued in senior grades, when the
real and then the complex numbers are introduced. Finally the secondary school
graduate has a stereotype formed. It is: ”everything can be measured by a number or a
corresponding widening of a number notion”.
European Research in Mathematics Education I.II: Group 6 191

So as to destroy the stereotype mentioned above we make pupils know permutations
and work at them. It is performed in the form of a game in the part ”For those who want
to have secret correspondence with friends” of the book ”Making Friends with
Algebra” in which the permutations are used to ”measure” the process of coding and
decoding texts. For instance, let us take the word ”timer”. We denote all letters in this
word by numbers of positions which they occupy in the word. If we want to code this
word, then we choose any order of numbers from 1 to 5, for example, 2, 5, 1, 3, 4 and
then we have the figure with two rows, which is named as permutation. This figure has
the meaning: in upper row there is initial order of letters in the word (i.e. 1, 2, 3, 4, 5), in
lower row there is the order of letters in coding word (i.e. in our case 2, 5, 1, 3, 4) and
numbers from upper row transfer to corresponding numbers from lower row (i.e. in our
case 1 transfers to 2, 2 transfers to 5 and so on). The permutation which gives the coding
word is named a coder and as result we have the coding word ”mteri”.

If we think the coder insufficiently secures the secret information, we can repeat the
same procedure with other coder to coding word and as result we have twice coding
word. For instance, let us take the other coder as permutation which upper row is 1, 2, 3,
4, 5 and which lower row is 1, 5, 2, 3, 4. If we apply this permutation as coder to coding
word ”mteri”, then we receive twice coding word ”merit”. What is the coder which
transfers initial word ”timer” to twice coding word ”merit”? In the first coder we have 1
transfers to 2, in the second coder we have 2 transfers to 5, then as result 1 transfers to 5
and so on for 2, 3, 4, 5. Finally we have twice coder as permutation, which upper row is
1, 2, 3, 4, 5 and which lower row is 5, 4, 1, 2, 3 (i.e. 1 transfers to 5, 2 transfers to 4, 3
transfers to 1 and so on). The twice coder is considered as the result of product
operation for two permutations.

So we are coming to definition of algebraic operation over permutations. Basic
properties of this operation also have real world interpretation. For instance, the
permutation with identical rows is named as identity permutation and corresponds to
transferring information without coding. Let us give a permutation, which is a coder,
and we change upper row with lower row in it, then resulting permutation performs
decoding and is named as decoder. After applying the coder to initial word and then
applying the decoder to coding word we receive the initial word. So the coder is inverse
European Research in Mathematics Education I.II: Group 6 192

element to the decoder relative to algebraic operation of permutation product. Also
such properties of this operation as associativeness and noncommutativeness are given.

At last pupils appear to be intellectually prepared to perceive the ideas of
classification of algebraic and numeric structures.

We invite pupils to meet the planet ”Quarta” 3 under conditions of this fantastic
planet children have to work with the number like objects in the situation when it is
impossible to create the fractions field of some polynomial ring. Pupils compare what
is common for numeric systems and number like objects. Finally they have an
opportunity to go over to the notions about group, ring and field.

7. Computer Visualization of Algebraic Structures

So we come to the hypothesis : forming Algebra understanding passes through such
stages as visual images- intuitive ideas-abstract notions- logical reasoning. This
process needs computer support. It is given by computer program ”Visual Algebra”,
which was worked out under my direction by Tomsk University computer science
undergraduate A.Shprenger. The idea of computer visualization of algebraic structures
is an application of discrete functions for constructing three-dimensional and two-
dimensional images of a given algebraic structure. Moreover they can manipulate by
these visual images, i. e. visual images are dynamic. Therefore the same visual images
from ”Visual Algebra” may give different intuitive ideas to students depending on the
state of their individual mental experience and psychological individual learner’s style.
Then the transition from visual images to intuitive ideas is an interesting research area
for a teacher. For instance pupils may set different algebraic operations independently
and to explore their properties. It is possible to see two – and three-dimensional images
of the assigned algebraic operations and to compare corresponding visual images.

This way is suitable for mixed-abilities classes. Every student realizes the transition
from visual images to intuitive ideas by the way which is most suitable for him. The key
problem – how this diversity of intuitive ideas will come to the correct abstract logical
reasoning? It requires high level of competence from a teacher. The most typical form
European Research in Mathematics Education I.II: Group 6 193

of intuitive idea is the recognition of visual images, for example, to recognize visually
different properties of algebraic structures.

As an education example we regard two groups: cyclic group of order 4, which is
looked upon as a subgroup of symmetric group of the fifth degree, generated by a cycle
of length 4 and Klein’s quarter group, regarded as a subgroup of symmetric group of the
fifth degree, generated by two cycles of length 2. Comparison of visual images of these
groups leads students to the hypothesis that these groups are not isomorphic. The next
step is logic basing of this hypothesis by comparison of elements orders in these groups.
At the same time the consideration of visual images for permutations cyclic group of
order 4 and residue classes additive group by modulo 4 leads to the hypothesis that
these groups are isomorphic. Logic basis of the given hypothesis may be found by
construction of isomorphism among the pointed out groups as concrete map.

Finally I invite all researchers, educators and everyone else who are interested in
”Visual Algebra” contact me for cooperation in exploring the possibilities, which are
given by ”Visual Algebra”. You are welcome to the beautiful world of visual algebra!

8. References
Weyl, A. (1995): ‘Topology and abstract algebra as two roads of mathematical comprehension’,
Amer. Math. Monthly, 102, 1-5.
Shafarevich, I. R. (1986): ‘Algebra-1’. Modern problems of mathematics. Fundamental directions,
vol.11, VINITI, Moskow.

1. The series of textbooks for children of 10-12 in MPI-project includes the following
i) “Decimal fractions in the Moominhouse”; ii) “Positive and Negative Numbers in
Pinocchio’s Theatre”; iii) “The Case of Divisibility and Other Stories”; iv) “About
Helen the Beautiful and Ivan the Prince and Fractions”.
European Research in Mathematics Education I.II: Group 6 194

2. The systematic course of algebra includes the following textbooks: “Marking
friends with Algebra”, “Identities”, “Algebraic fractions”, “The Book about the
Roots from which New Numbers Grow”, “Quadratic Equations”, “Inequalities”,
“Equation Systems”, “A Fairy-Tale about the Sleeping Beauty or a Function”,
“Quadratic Function”, “Sequences”.
3. A part “Ksysha on Quarta” of the book “Algebraic fractions” is devoted to this
problem (8 grade ,14 years old).
European Research in Mathematics Education I.II: Group 6 195

Éva Szeredi, Judit Török
Teacher Training College of ELTE, Budapest

Abstract: The aim of this paper is to show how some research tools coming from
different disciplines and theoretical frameworks were used to explore students’
difficulties in a rather complex algebra task and the differences between two different
groups of students, in order to produce interpretative hypotheses about emerging
Keywords: tools, comparison, performances, algebra.

1. Introduction

A research study was performed by us at the Teacher Training College of ELTE
(Budapest) about the use of a set of tools (derived from different sources) to analyse
students’ performances and interpret their difficulties in algebra tasks. These tools
concerned both a-priori and a-posteriori analysis, and included peculiar keys to analyse
processes (anticipation, exploration, cognitive unity between conjecturing and proving,
etc.). In this paper we will exploit some of these tools (which will be described in the
a-priori analysis section - see 2.2. and 2.3.) to analyse and compare Italian and
Hungarian university students’ performances in the following task: “Generalize the
property: ”The sum of two consecutive odd numbers is divisible by four”, and prove the
generalized property”. The aim of the reported study was to produce interpretative
hypotheses about emerging phenomena.

This report is based on the protocols of 33 Italian and 41 Hungarian students who
worked on the task for approximately 90 minutes. The Italian protocols were made by
three groups of 4th year mathematics students at the Genoa University. After
completing their 3-year course on mathematics, they chose to specialize on
mathematics teaching. The Hungarian protocols came from two groups of 2nd year
European Research in Mathematics Education I.II: Group 6 196

students, one group of 3rd year students and one group of 2nd year evening-course
students. All of them are trained for teaching mathematics (for children age 10-14)
from the beginning of their studies at our college. All of them had already finished one
semester on number theory before the task was given to them (no course of this kind
was attended by Italian students). The task was quite unusual for both the Italian and the
Hungarian students, it seemed to be appropriate for collecting data about the students’
ability in making generalizations, conjectures and proofs.

2. A Priori Analysis of the Task

2.1 Some Possible Solutions

There are several ways for generalizing the original statement.
• We expected most the following:
”The sum of 2m consecutive odd numbers is divisible by 4m.
Different proofs are possible, for example:
Proof 1: The sum of 2m consecutive odd numbers is:
S = 2k–2m+1 + 2k–2m+3 + 2k–2m+5 + ...+ 2k+2m–5+ 2k+2m–3 + 2k+2m–1
S = 2m .2k = 4mk ; finally we get: 4m |S.
Proof 2: Starting with 2k+1, the nth odd number is 2k+2n–1. The sum of n
consecutive odd numbers is S = 2k+1 + 2k+3 + 2k+5 + ... + 2k+2n–1.
Applying the formula for the sum of an arithmetical sequence we get:
S = n(2k+1+2k+2n–1)/2 = n(2k+n). If n is even, e.g. n = 2m,
then S = 2m(2k+2m) = 4m(k+m), which is clearly divisible by 4m.
• There are some ”weaker” generalizations as well, for example:
”The sum of an even number of consecutive odd numbers is divisible by 4.”
• There are also some other ways of generalization (see 3.2. C).

2.2 Reference Knowledge

i) The task needs some basic knowledge of arithmetics and algebra like:
• concept of divisibility, consecutive odd/even numbers. These concepts are
obviously necessary to understand the task;
European Research in Mathematics Education I.II: Group 6 197

• algebraic representation of divisibility and of consecutive odd/even numbers;
• ability to express the nth term of certain series of numbers;
• appropriate use of parameters.
These skills refer to the ability and confidence at the algebraic level. They are
basically necessary if someone wants to give a proof for a meaningful general
ii) Mastery of some formulae, results in elementary number theory and methods
ready-to-use as a tools, are necessary or may be very useful:
• transformation of algebraic expressions;
• the formula for the sum of the first n odd or at least even numbers or the general
formula of the sum of an arithmetic progression or the ability to construct or
re-construct such a formula quickly;
• induction method.
Transformations of algebraic expressions are necessary in proving. The knowledge
of a useful formula can be a great help at the proving stage but if missing, it can be
replaced (for example by constructing and proving the formula). Induction can be
useful at that case.
iii) The necessary metaknowledge about:
• generalization: some understanding of the meaning of generalization is strictly
necessary at least on an intuitive level;
• the structure of the statement of a theorem (hypotheses, thesis). It is needed to
write down the statement in an appropriate way and prepare the proving stage;
• proof: the knowledge of the requirements needed for mathematical proof (as a
deductive chain, etc.) is necessary not only to write down the final product, but
also to construct the proof.

2.3 Processes

Producing the conjecture and arguments supporting it and constructing the proof needs:

A) Anticipation
It is a key process, which enters both at the semantic level (planning appropriate
European Research in Mathematics Education I.II: Group 6 198

numerical experiments) and at the algebraic formalism level (planning appropriate
transformations). (see Piaget, 1967; Boero, 1997).

B) Exploration
(see Polya,1962; 1973; Simon, 1996; Boero et al., 1996).

• Checking the original statement (through arithmetical examples or by algebraic
• looking for variables to be left free in order to find a way for generalization;
• arithmetic or algebraic exploration of the situation:
– looking for numerical regularities
– exploring general formulae in order to select a regularity expressed in
algebraic terms
– testing the idea of a conjecture in specific cases

These are constitutive elements of the complex processes explicitly demanded by
the task.

C) Producing the statement
• Elaborating ideas (guesses for a stable statement, arguments supporting it) taken
from experiments and explorations;
• Constructing and expressing the conjecture in words and/or by formula.

D) Proving
• Detecting and exploring appropriate arguments from the preceding phases:
formalization of arguments, interpretation of algebraic expressions (Arzarello et
al., 1995), possibly establishing a ”cognitive unity” with the production of the
conjecture (see Mariotti et al., 1997; Garuti et al., 1998);
• producing new arguments, if necessary. It may require transition from one frame
(e.g. arithmetics) to another (e.g. algebra) (see Arzarello et al., 1995);
• connecting the appropriate arguments in a deductive chain.
European Research in Mathematics Education I.II: Group 6 199

The passage from algebraic formalism to the semantics and vice-versa
(”interpretation” and ”formalization”) are key elements in proving and in some cases in
the creation of the conjecture.

2.4 Difficulties

• Difficulties depending on the subject’s capacities and knowledge may arise from
lack of the anticipatory thinking and/or lack of the adequate arguments.
• Difficulties depending on specific aspects of the strategy may arise from:
– breaking the continuity between producing the conjecture and constructing
the proof (see Garuti et al., 1998; Mariotti et al., 1997) (this breaking may
concern the kind of exploration, formalism, frames and reference culture);
– lack of flexible usage of formalism (it means: adapting the formalism to the
situation) (see Duval, 1995).
• Difficulties with ”metaconcepts” (about generalization, proof, etc.)
• Difficulties related to the didactical contract (Brousseau, 1986): under an
unusual task like this, students try to guess the teacher’s expectations deriving
from their past experiences, but those experiences are far from the present task,
so their effort may be oriented in an inappropriate direction.

2.5 A Priori Expectations

Our expectations were based both on the preceding a-priori analysis and our knowledge
about students’ background and performances in other algebraic tasks.

We expected the majority of the students to start by proving the original statement;
than explore the problem arithmetically and/or algebraically, by increasing the number
of terms, proving some specific cases and a minority to try to expand the problem in
other ways. We expected that the notion of generalization would have not been entirely
clear for the students, that their conjectures would have been a mixture of proper
generalizations (weaker or stronger) of the original statement and of other statements
somehow connected with the initial one. We expected that a good percentage of the
European Research in Mathematics Education I.II: Group 6 200

students would have been able to prove the produced conjecture.

3. Analysis of Performances

3.1 How the Students Exploited or not Reference Knowledge

i) Every student was able to understand the original statement. We found that none of
the students had problems with arithmetic elements like divisibility, odd or even
consecutive numbers, and all of them were able to find an algebraic representation
for two (or a few more) consecutive odd/even numbers. More problems emerged in
the use of formalism, especially as mastery of parameters is concerned (to
accomplish less standard tasks).
ii) The success of finding a solution depended strongly on the confidence in the field of
algebra, e.g. on the ability to find a formula for the general term.
Some students had difficulties with the transformation of a formula when they tried
to use the ? sign. Some students tried to use induction for proving divisibility (it may
depend on a lack of anticipation at the usage of this method). Those who were
familiar with the sum of the first n odd or even numbers were in a much easier
situation then those who were not. In some cases, the use of induction could have
been a help at that stage.
Being able or not to interpret an algebraic formula was a very important factor of
success in finding and proving a meaningful conjecture; in many cases failure or
success depended on it.
iii) There was a lot of confusion about the meaning of generalization. As we will show
later, among the conjectures produced there were quite a few trivial generalizations
and there were a lot which were true (or false) statements connected with the initial
problem but not generalizations of that.
There were great differences at the success in proving the conjectures, depending on
metaknowledge about proof. Indeed, while there were quite a few protocols in
which we could not find any serious attempt for proving, there were others who
became stuck at the beginning because of confusion about the nature of proof. There
were also some students showing consciousness about some aspects of proof
European Research in Mathematics Education I.II: Group 6 201

(generality, exploitation of hypotheses ...) but unable to detect a missing link in the
deductive chain.
iv) The didactical contract seemed to have a significant influence on the products
especially in connection with the generalization process and proving process.

3.2 Processes

A) Anticipation
Anticipation is an important factor both in the exploration and the proving phases. We
will refer to its role there.

B) Exploration
Almost all of the students started their work with proving the original statement. Most
of them constructed an algebraic proof but they were not equally confident in the
algebraic representation of two consecutive numbers and of the sum of them. Some of
them used the most proper parameters and representations quite easy to interpret,
like:”2k–1 + 2k+1 = 4k” or ”2k+1 + 2k+3 = 4k+4”. Some others used more complicated
systems of parameters, representations with more then one step, sometimes not
covering all the cases, like:

”a + a+2 = 2(a+1) where a is an odd number” or
”4k+1 +2m+1 = 2(2k+m+1) where m is an odd number”.

After (or without) proving the original statement almost every student started –
consciously or instinctively, systematically or at random – to look for an explicit or
implicit parameter which can be changed. All of them made – either arithmetical or
algebraic – experiments with statements of the form ”The sum of any number of
something is divisible by something”. They tried several possibilities.

Most of the students started to experiment with:
• the sum of 3, 4, ... consecutive odd numbers (instead of 2)
• the sum of any two odd numbers (instead of consecutive)
• the sum of two consecutive whole or even numbers (instead of odd).
European Research in Mathematics Education I.II: Group 6 202

Some of them changed more than one parameter of the original statement, for
example experimented the sum of more than two consecutive whole numbers.

Some of them experimented with more hidden parameters, for example with
(2k+1)n + (2k+3)n.

The students either tried to find cases where their sum is divisible by 4, or tried to
change 4 for an other divisor.

The exploration of two consecutive whole or even numbers does not lead to a
generalization. Those who chose this direction– either because of the lack of
anticipation or because being very systematic while working on the concept of
generalization– got stuck or changed for an other direction.

Many of those who insisted on the divisibility by 4 were not able to find more than
weak generalizations. On the other hand, those who changed too many parameters at
the same time during their experiments easily became lost while collecting lots of
statements similar to the original one, or got (sometimes quite meaningful) general
statements which were not generalizations of the original one.

As we could see, anticipation played a very important role at choosing a proper
parameter to change.

Experimenting with 3, 4, 5, 6, ... consecutive odd numbers turned out to be the most
promising direction. During the exploration the students worked on three different level
of abstraction:

I. Level of number examples.
II. Level of statements we can obtain from the original one by changing one given
value in it for another value. For example statements like ”The sum of three
consecutive odd numbers is divisible by 3”. These statements have the same number
of variables as the original statement and all of them can be fulfilled by infinitely
many number examples. The students could get them either from arithmetic or
algebraic experiments but their verification can not be done by using number
examples only.
European Research in Mathematics Education I.II: Group 6 203

III.Level of statements we can obtain from the original one by changing a given
value(s) for variable(s). For example statements like ”The sum of 2k consecutive
odd numbers is divisible by 2k+1 ”. A statement of this level was a common
generalization of a subset of second level statements. The task was to find a
statement of this sort where the subset contains the original statement.

The results of the students strongly depended on the level they are familiar with and
on their ability of moving from one level to an other.

Those students who had very poor algebraic tools were confident only at the first
level. Those of them who experimented only with number examples were able with a
few exceptions to construct a good (in some cases even a strong) conjecture expressed
in words but they had no tools for proving. Their success in finding a good conjecture
depended on their ability of finding some order in the experiments.

In spite of being uncertain of the algebraic representation even of consecutive odd
numbers, some of them tried to explore the situation algebraically. They often got stuck
at the different parametrizations of either the original or an other second level

Those students who were able to move on to work algebraically with second level
statements mostly chose sooner or later to experiment with the sum of 3, 4, 5, ...
consecutive odd numbers. They realized quite soon that the sum of 3 (5, 7,...)
consecutive odd numbers is not divisible by 4, and they reacted to this fact in different
ways. Some of them became confused, changed the direction by chance without
anticipation, or changed the divisor too, or changed too many parameters. Some of
them tried to construct a statement which is true for all the cases, like ”The sum of n
consecutive odd numbers is divisible by n” or ”The sum of n consecutive odd numbers
is divisible by their mean”. They forgot that the result should not be merely the common
generalization of their second level statements but of the original statement as well.
Many of them noticed that separating the cases (odd number of members and even
number of members) would help and they got a good conjecture. Those again, who
rigidly insisted on the divisibility by 4, got only the weak generalization.
European Research in Mathematics Education I.II: Group 6 204

C) Producing the statement
Some of the students did not give any statement as a generalization or considered a
second algebraic representation of the original statement as a generalization. For
example: ”Not only 2k+1 + 2k+3 = 4(k+1) is divisible by 4, but if p is odd, then 2k+p +
2k+p+2 = 2(2k+p+1) is also divisible by 4.”

Some of them gave only statements of the second level, like ”The sum of three
consecutive odd number is divisible by 3.” It is not always clear from the protocols if
they considered these statements as generalizations or just as results of experiments.

Some of them gave third level statements which were generalizations for their
second level experiences but not for the original statement. For example: ”The sum of n
consecutive odd numbers is divisible by n.” Again, it is not always clear from the
protocols if they considered these statements as generalizations or just as results of

Some of them gave a proper generalization but only a trivial one, like ”If two odd
numbers have different remainders by 4, then their sum is divisible by 4.”

Many of them constructed meaningful generalizations like:
• ”The sum of 2n consecutive odd numbers is divisible by 2n+1.”
• ”The sum of 2k consecutive odd numbers is divisible by 4k.”
• ”If a and b are consecutive odd numbers and n is odd, then an+bn is divisible by
• ”If we take n2 consecutive numbers and then leave out those which are divisible
by n, then the sum of the remaining n2-n numbers is divisible by n2.”

Algebraic tools were helpful but not necessary for constructing even a meaningful
conjecture. Those students who were not successful at getting one either lacked
anticipation and were not flexible when their experiences contradicted to their
expectations, or were not able to keep a systematic order while making experiments, or
were too uncertain about the concept of generalization.

Many of the students gave more than one (not always proper) generalizations. The
European Research in Mathematics Education I.II: Group 6 205

quality and the amount of their statements also depends on what they thought the
expectations were

D) Proving
While exploration was possible on the first two levels, to prove it is unavoidable to
move to the third level. Now we deal only with those who arrived at a meaningful
conjecture. The proving phase was strongly interconnected with the exploration phase
in the students’ works. There were great differences between the students in terms of
the arguments collected before they arrived at the proving stage. The two extremes
• those who created a conjecture in words based only on numerical exploration;
• those who produced a correct closed formula of the sum of an even number of
consecutive odd numbers, and their conjecture was only the interpretation of that

With a few exceptions the students who tried to prove a general statement dealt with
statements about the sum of consecutive odd numbers. The proof needs the ability to
formalize the sum of n terms of a series and to give a formula for the general term (i.e.
the ability to formalize their conjecture). To cope with these difficulties, the arguments
one collected so far came very useful (”cognitive unity”, see Mariotti et al., 1997).
There were basically two ways of using the arguments collected during the exploration:
• The arguments were used immediately when they appeared, as a part of a
continuous process. They served in these cases like the steps of the stairs for their
• The arguments already left behind were taken out again when the progress met
an obstacle or when there was a gap in the deductive chain. They served in these
cases either as starting points to avoid the obstacles or element to bridge the gap.

These ways resulted in a continuity in all the processes. Breaking this continuity was
quite rare, but caused disaster in most of the cases.

To go on with the proving process one had to be able to find a ”closed” formula for
the sum and to interpret it.
European Research in Mathematics Education I.II: Group 6 206

Finding a closed formula was very difficult for many students.

Some of them tried to use the S sign, but it did not help. The formal usage of
algebraic symbols (without the clear understanding of their meaning) broke the
continuity in many cases causing confusion rather than helping the students as the
following example shows.

”(2n+1) + (2n–1) = 4n OK!
(2n+1) + (2n–1) + (2n+3) + (2n–3) = 8n still divisible by 4, but by 8 as well!!
Similarly, for the sum of consecutive numbers, we will always get 2n + ... + 2n this will
always be divisible by 4, if I take even number of consecutives. Moreover if I take 2k
consecutives, their sum is divisible by 4k.
I prove what I have found: (2n+1) + (2n–1) + ... + (2n+p) + (2n–p) =
S p=1 to k (2n+p) + (2n–p) = S p=1 to k 4n = 4n but it can not be 4n!!”
(At this point he starts again without any more results.)

Some others tried proving by induction but it was evidently hopeless in this
situation. These cases clearly show the lack of anticipation in the application of these
techniques. Some students transformed the sum into another, perhaps a simpler but not
a closed one. For example:

”I translate the fact of having m (even) consecutive odd numbers:
2n+1, 2n+3, ..., 2n+(2m–1). Their sum is : 2nm + (1+3+...+(2m–1)).
2nm is divisible by 2m, so it is enough to show that 1 + 3 + ... + (2m–1) is divisible by 2m.
I know that the sum of two consecutive odd numbers is divisible by 4:

1+3 + 5+7 + … + (2m–3)+(2m–1) =
4•s 4•t 4•k
4•1 4•2 4 • (m–l)
= 4(1 + 3 + ...+ (m–1)) where the sum has m/2 members.”
(At this point he thinks that the last sum has odd number of members and tries to prove –by
a hopeless induction– that the sum is divisible by the number of members).

To find a closed formula was a new ”problem in the problem” and they could start a
new exploration-hypothesis-proof cycle (induction could have been helpful for proving
European Research in Mathematics Education I.II: Group 6 207

the validity of a formula for the sum). This path could have been successful but no one
could complete it in the given time.

Producing a complete proof demanded good anticipation, enough arguments from
the exploration, some flexibility, confidence in the necessary algebra, and clear
understanding of what proving means.

4. Comparison Between Hungarian and Italian Students

4.1 Findings

We have found many similarities between the Italian and Hungarian protocols, all types
of difficulties and typical reasoning occurred in both cases. Now we will concentrate on
the differences.

The differences we found fall into three main types: understanding the task; the use
of algebra; consciousness about proving.

The Italian students were more aware of the concept of generalization. Some of
them explicitly stated what generalization means (while none of the Hungarians) and
quite systematically tried to find a proper free variable. Even if some of them could not
produce a meaningful general statement, none of them moved away from the task. The
Hungarian students had only an intuitive concept of generalization and many of them
produced various general statements which were not necessarily generalizations of the
original one. Some of them moved away quite far from the task.

Italian students had more problems with the necessary algebra and basic results of
elementary number theory than the Hungarian ones. Some of them tried to use algebra
on a level which they were not familiar with. The Hungarian students seemed to be
more comfortable with using algebra and basic results of elementary number theory as

Italian students felt more obliged to prove. Even if they were not able to complete a
proof they were more keen on trying it. Some of the Hungarian students were more
European Research in Mathematics Education I.II: Group 6 208

successful at proving (due to their algebra and number theory knowledge) but some of
them did not even try to prove, and some were satisfied with a convincing argument.

4.2 Possible Interpretations

Although taken from small samples occasionally collected and not representative of
Hungarian and Italian students, the preceding quantitative and qualitative data show the
existence of important differences between the two groups of students probably
depending on their pre-university and university background. Some of the research
tools considered in this paper can provide us with interpretative hypotheses about
phenomena emerging from our exploratory study. Further, more extensive and
statistically significant comparisons should be performed in order to test these
hypotheses, keeping into account that the complexity of the task and of the students’
background involved in it represents a tremendous challenge for a serious comparative

Above the obvious cultural/educational differences we feel that the main reasons
behind the findings above are the different curricula and the different didactical
contracts. While in Hungary all the secondary school students learn and have to solve a
lot of exercises about the sum of progressions and traditionally there is a great emphasis
on number theory, we think that in Italy students are trained to be more conscious about
the most important metaconcepts of mathematics. The produced protocols helped us to
realize the importance of the didactical contract and to reveal some elements of it.
There were significant differences even between the Hungarian groups taught by
different lecturers. Hungarian students are not expected and have very little practice in
reflecting to (and controlling) their thinking. They would not be able to put down
private comments about their reasoning and difficulties even if they were asked for it.
The fact that many Hungarian students considered the task as an open investigation
problem was partly due to their previous education, but, as the differences between the
Hungarian groups show, to the personal didactical contracts too.
European Research in Mathematics Education I.II: Group 6 209

5. Conclusion

The performed study confirms the need for tools of different sources (epistemology,
psychology, didactics) in order to analyse this kind of task.

Epistemology and psychology tools seem to be sufficient to analyse individual
behaviours; but when comparison between different groups is performed, the didactical
contract (typical tool from didactics of mathematics) is needed in order to interpret
differences which seem to be inaccessible to other interpretations.

Finally we would like to point out the importance of students’ metaknowledge
(about generalizing and proving) in open tasks like that considered in this paper.

6. References
Arzarello, F.; Bazzini, L. and Chiappini, G.P. (1995): The Construction of Algebraic Knowledge,
Proc. of PME-XIX, Recife, vol. 1, pp. 119-134.
Boero, P.; Garuti, R. and Mariotti, M.A. (1996): Some Dynamic Mental Processes Underlying
Producing and Proving Conjectures, Proc. of PME-XX, Valencia, vol. 2, pp.121-128.
Boero, P. (1997): Transformation and Anticipation as Key Processes in Algebraic Problem Solving,
in R.Sutherland (ed), Algebraic processes and structures, Dordrecht: Kluwer A.P. (in press).
Brousseau, G. (1986): Fondements et méthodes de la didactique, Recherches en Didactique des
Mathematiques, 7, 33-116.
Duval, R. (1995): Sémiosis et pensée humaine, Peter Lang, Bern.
Garuti, R.; Boero, P. and Lemut, E.: Cognitive Unity of Theorems and Difficulty of Proof, Proc. of
PME-XXII, Stellenbosch, Vol. 2, pp. 345 - 352.
Mariotti, M.A.; Bartolini Bussi, M.; Boero, P.; Ferri, F. and Garuti, R. (1997): Approaching Geometry
Theorems in Contexts: From History and Epistemology to Cognition, Proc. of PME-XXI, Lahti,
vol. 1, pp. 180-195.
Piaget, J. (1967): The child’s conception of space, New York: W.W. Norton & C.
Polya, G. (1962): Mathematical Discovery, John Wiley, New York.
Polya, G. (1973): How to solve it? Princeton, NJ: Princeton University Press.
Simon, M. (1996): Beyond Inductive and Deductive Reasoning: The Search for a Sense of Knowing,
Educational Studies in Mathematics, 30, 197-210.

European Research in Mathematics Education I.II: Group 7 211

Milan Hejny1, Christine Shiu2, Juan Diaz Godino3, Herman Maier4
Faculty of Education, Charles University, Prague, Czech Republic
The Open University, UK
University of Granada, Spain
Universitaet Regensburg, Germany

(Elaborated by Christine Shiu)

The seven papers which comprised the starting point for our discussion and are
published above seemed to fall naturally into three types: two dealt with global theories
of mathematics education (Rouchier, Godino & Batanero), three were empirical studies
which addressed specific questions in mathematics education (Bagni, D’Amore &
Maier, Stein) and the remaining two presented accounts of elaborated methodologies
which have been devised and developed for specific purposes in research into
mathematics education (Marí, Stehlíková).

It was agreed in the group that we would allow one session of 45-60 minutes in
length (known as an A-type session) for each paper to be discussed intensively and that
such sessions would be interspersed with longer ones (90 minutes, known as B-type
sessions) in which general issues emerging from the particular cases would be
examined. In A-type sessions the author-presenter gave a brief synthesis of the paper
and added any complementary information which was deemed appropriate. The chair
then moved the group from questions of clarification into the identification of the issues
European Research in Mathematics Education I.II: Group 7 212

The pattern of the sessions was as follows:

Day 1
Session A1 André Rouchier, chaired by Klaus Hasemann
Session A2 Juan Godino, chaired by Leo Rogers
Session A3 Giorgio Bagni, chaired by Martin Stein
Session B1 general discussion, chaired by Hermann Maier

Day 2
Session A4 Hermann Maier, chaired by Jeremy Kilpatrick
Session A5 Martin Stein, chaired by André Rouchier
Session A6 José Marí, chaired by Nada Stehlíková
Session B2 general discussion, chaired by Juan Godino

Day 3
Session A7 Nada Stehlíková, chaired by Gunnar Gjone
Session B3 general discussion, chaired by Christine Shiu

The resultant general discussion involved us in a consideration of the nature, roles
and functions of theories in mathematics education and how these related to specific
questions in mathematics education, and how both theories and specific questions
shaped and were shaped by the methods used to investigate the questions. What
emerged was a triad of theory, research questions and methods, which when developed
and used in a particular context, could viewed as a research paradigm. It was noticeable
that specific research paradigms seemed often to be particular to the cultural or national
context in which they were developed.

It was soon apparent that interpretations of the meanings of many words which we
were using differed among the members of the group. These differences too might be
particular to their cultural or national context. Elucidation of differences framed the
discussion as they were explored through a series of questions. The questions
themselves were in their turn scrutinised and refined as the sessions progressed, and the
following summarises the progress and scope of the discussion.
European Research in Mathematics Education I.II: Group 7 213

1. Theories in and of Mathematics Education

1.1 What is a Theory and How Do Theories Arise?

We took as a working definition that a theory is a “framework of concepts which show
how things work”.

Theories in and of mathematics education can be developed within mathematics
education but often, quite properly borrow from and particularise theories developed in
other disciplines such as psychology, sociology, anthropology, semiotics etc. Theories
can therefore be imported from other fields of knowledge (and used unchanged or
adapted to a certain degree) or be the results of previous research in mathematics
education or “inventions” on the basis of an implicit and holistic pre-understanding of
the field of research.

1.2 What Is the Role and Function of Theory in the Research Process?

Theory is not imperialism but simplification. Theory functions in the research process
as a means of reducing and controlling the variables that have to be taken into account
when the researcher is studying, for example, a didactical process. Mathematics
education is a domain in which there are a large number of salient variables.
Researchers’ patterns of interaction with the reality studied necessarily involve
reducing and controlling some of those variables, in order to allow them to attend to
those they investigate. Theory is thus one means (the most powerful) we have of
making an a priori analysis of the variables inherent in a situation, and hence of
allowing us to make rational and defensible choices about what variables to control.

For example, consider the work of the researcher who is creating a classroom
situation with the help of didactical engineering. The a priori analysis is the study of all
possible events (from a cognitive and mathematical point of view) in the development
of that situation according to the specific mathematics problems and to the other
components of the didactical setting. It corresponds to the “task analysis” of
psychologists. This analysis then aids the interpretation of the actual events, so helping
our understanding of the process under study.
European Research in Mathematics Education I.II: Group 7 214

Theory is also a means of generating questions and problems that are practical,
operational, open to empirical study. The same a priori analysis of the situation often
identifies important phenomena susceptible to investigation. It shapes the view of the
reality to be investigated and opens a particular perspective on this reality, hence
raising and precising questions and equipping the researcher with a language to
formulate these questions. Finally it guides the data collection and the data analysis and
is thus a determinant of the research methods to be used.

Sometimes it is appropriate to start with an empirical exploration of the research
field without an explicit theoretical model. This happens when the researcher wants to
remain open for a deep and adequate understanding of the reality. Reflection on the
analysed data can result in a description or an explanatory theory. Theory generated this
way is known as “grounded theory”, comprising a posteriori accounts emerging from
observed data. This notion seems to be consistent with the working definition given
above. However it is essential that the process is completed, and that researchers’
claims to be developing grounded theory are substantiated in their reports with specific
accounts of the emergent theory being provided.

Some more specific functions of theory might actually demand the production of
different kinds of theory. Four possibilities are:

• descriptive theories which give an account of what has happened, describe what
is the case;
• explanatory theories which seek to explain why or how something happened;
• predictive theories which predict what will happen in given conditions;
• action theories which guide action by identifying what can be done in given

Identifying such possibilities raises a further question.
European Research in Mathematics Education I.II: Group 7 215

1.3 Is It Feasible and Useful to Distinguish Different Kinds of Theory?

Perhaps partial theories might better be designated “models” reserving the word
“theory” for the larger frameworks of concepts which offer more global accounts of
how things work. It may be noted that the field of mathematics education can be broken
down into different kinds of elements and that different elements would demand
different global theories. They would also lead to different research questions and
different methods of investigation. In the two examples in this section of the
proceedings the elements are “didactical situations”(Rouchier) and “meanings of
mathematical objects” (Godino).

1.4 How Can Theories in Mathematics Education Be Evaluated?

Mathematics education is not mathematics, nor is it a science. Its theories cannot be
proved by an “A implies B” chain of logical reasoning. Nor can we look for the
Popperian notion of falsification by a crucial experiment. Rather we must look for
self-consistent narratives in terms of identified elements within mathematics education
which are sufficient and are illuminating in accounting for the observed phenomena.
The criteria by which theories in mathematics education are judged are adequacy and
usefulness. Evaluation is therefore essentially pragmatic.

Published theories can also be subjected to external evaluation through data
generated independently of the theory. Can possessors of data apply a theory so as to
see their observations in a new way? Can generators of theory interpret external data in
terms of existing formulations? For example, what light can be thrown on Bagni’s
teaching experiment by the theory of didactical situations or by the theory of meanings
of mathematical objects? The focus of the paper attends more closely to the elements of
the latter theory. A relationship between the introduction of a group concept by means
of an historical example and an earlier investigation of the meaning of mathematical
objects could be posited. In particular, the introduction of the group structure could be
read as “a human activity involving the solution of social-shared problem-situations”
leading to the creation of “a symbolic language in which problem-situations and their
solutions are expressed”. (Godino & Batanero 1998, p. 179).
European Research in Mathematics Education I.II: Group 7 216

By such tests theories are more likely to be modified or extended rather than verified or
falsified in their entirety. They may also be extended, modified or changed through
theoretical reflection. Experience suggests that major paradigm shifts occur when a
theory is found inadequate.

2. Conducting Empirical Research

2.1 What Can Be Learned from Empirical Research?

Empirical research can be carried out in a number of physical and social contexts. In our
three examples Bagni reported a teaching experiment carried out in regular classrooms.
The textual eigenproductions (TEPs) discussed by Maier were created in classrooms
but the data included teachers’ responses to these TEPs outside a classroom setting. In
Stein’s study the social unit was a pair of students working on a task rather than the
whole class. The choices implied by these contexts derived from the phenomenon to be
researched and the particular questions to be addressed. In all cases the description of
the conditions of the study is an essential tool in allowing the reader to interpret the

It was noted that over the history of research in mathematics education, which was
seen as a twentieth century phenomenon, there has been a marked shift from
quantitative to qualitative methods and approaches. This probably reflects a view that it
is the exceptional case which challenges our perceptions of what is the case, and which
causes us to seek new interpretations and explanations. We therefore need
phenomenological accounts – thick descriptions – of individual cases. On the other
hand, a possible danger inherent in the exclusive pursuit of qualitative data is that we
may fail to establish the typical to which our special case is the exception. We may need
a broader picture, possibly established through quantitative approaches, to anchor our
new interpretations.
European Research in Mathematics Education I.II: Group 7 217

2.2 What Is the Relationship Between the Researcher and the

Another issue in empirical research is the effect of the observer on the observed,
especially but not exclusively, when the observer is a participant in the situation. In the
above discussion of theory it was noted that studying classroom processes becomes
possible when we are able to reduce the numbers of variables or to control some of
them. It follows that where an external researcher is studying the interactions in a
particular classroom the construction of the teaching to be done must be shared by the
teacher and the researcher. The technology of building situations (to research) is
supported by previous knowledge in the field (including in particular theoretical
knowledge). This has been described as ingenierie didactique or didactical engineering.

Action research is research carried out by practitioners on their own practice with a
view to changing and improving that practice. Two of the papers (D’Amore & Maier,
Stehlíková) described projects which involved teachers taking on aspects of the role of
researchers through a sharing of the methods of the projects. Discussion of this led to a
further question.

2.3 How Can Research Best Be Shared with Teachers?

There is a responsibility to report on research results in a format which is accessible to
teachers. However it is often more fruitful to share the methods of research – usually
qualitative methods – with teachers so that they can construct knowledge particular to
their own situation.

2.4 Developing Methodologies

What is a methodology?
As with theory there is a question about how global, how all-embracing a methodology
must be in order to be called a methodology. In both of the papers above (Marí,
Stehlíková) the methodology is detailed and elaborated. In both cases this elaboration
performs an integrating function – in the first a particular topic from the school
European Research in Mathematics Education I.II: Group 7 218

mathematics curriculum is chosen and a very thorough survey of existing research of
that topic is synthesised to give a basis for the study. In the second the project examines
a range of aspects of learning mathematics and seeks to integrate the work of many
researchers including teacher-researchers to produce the findings.

It was agreed that major projects need all-embracing methodologies. However
much valuable research is carried on a smaller scale and it was pointed out that without
such small scale studies there would be no existing research to survey. What is
important is that in reporting research the methods used and the reasons for the choice
of methods are reported as clearly as possible to allow others to follow up and build on
that work.

3. Conclusions

The conference structure – and themes – worked well for Group 7 so we relate our
conclusions to the themes of the conference.

3.1 Communication

All participants found great value in sharing accounts of research practice. We agreed
that an important aim (possibly the central aim) of our research is the improvement of
mathematics teaching. It follows that an important part of communication is
communicating with teachers of mathematics. What we communicate may be research
results, but it may be research methods, perhaps helping the reflective practitioner to
become a practitioner-researcher.

Whoever the audience of our research reports, in order to maximise communication
we need to be as clear as possible about the researchers’ underpinning theoretical
position, and about both the conditions and the methods of research.

One difficulty with communication with colleagues from different research
traditions lies in undeclared assumptions. Ambiguities and misunderstandings may
also arise from the different meanings and connotations of words which we use in
European Research in Mathematics Education I.II: Group 7 219


3.2 Cooperation

The spirit of cooperation was strong in the group.

Like all groups we started from the premise that papers would not be presented, but
would have been read by all. This proved to be a valid premise. Nevertheless we found
that our A-type sessions allowed a useful focus on individual papers from which
authors received feedback and ideas for improving final versions of papers and
participants gained deeper insight into what they had read.

3.3 Collaboration

As we worked there was a growing awareness of similarities and differences among our
perspectives. It seemed that research paradigms are in some ways particular to the
country in which they arise.

Our activities were essentially the first stage of a group project to improve
communication by clarifying commonly used terms such as: paradigm, theory,
didactics, mathematics.

4. References
Godino J.D. & Batanero C. (1998) ‘Clarifying the meaning of mathematical objects as a priority area
for research in mathematics education, in Sierpinska A. & Kilpatrick J. (1998) Mathematics
Education as a research domain: a search for identity, Kluwer, pages 177-195.
European Research in Mathematics Education I.II: Group 7 220

Giorgio T. Bagni
Nucleo di Ricerca in Didattica della Matematica, Department of Mathematics, University of Bologna, Piazza di Porta
san Donato 5, I- 40127 Bologna, Italy

Abstract: The effectiveness of the use of history of mathematics in mathematics
education is worthy of careful research. In this paper the introduction of the group
concept to an experimental sample of students aged 16-18 years by an historical
example drawn from Bombelli’s Algebra (1572) is described. A second sample of
students was given a parallel introduction through a Cayley table. Both groups were
asked the same test questions and their responses examined and compared.
Keywords: history, algebra, group.

1. Bombelli’s Algebra (1572) and Imaginary Numbers

Several authors have shown that the history of mathematics can be drawn on by
teachers in the presentation of many mathematical topics to the benefit of pupils. It
follows that research into the role of history of mathematics in teaching is legitimately
considered a part of research into mathematics education (many references can be
mentioned; for example: Jahnke, 1991, 1995 and 1996).

Of course we must consider the educational use of history of mathematics at
different levels, and these levels can lead to different educational interventions. For
example, according to the conception of the mathematics education as thought
transference, the main purpose of the educational research is improvement of teaching.
The presentation of mathematical topics using historical references is consistent with
this approach. Of course, the effectiveness of the historical introduction will be judged
with respect to pupils’ learning.
European Research in Mathematics Education I.II: Group 7 221

In this paper, we consider an important topic of the curriculum of both high school
(for students aged 16-18 years) and undergraduate mathematics, namely the group

Rafael Bombelli of Bologna (1526-1572) was the author of Algebra, published
twice, in 1572 and in 1579 (the dispute between G. Cardan and N. Fontana Tartaglia
about the resolution of cubic equations is well known; let us underline that Scipio Del
Ferro was remembered in Bombelli’s Algebra: manuscript B.1569, Archiginnasio
Library, Bologna).

Fig. 1: Bombelli’s Algebra (1572-1579)
European Research in Mathematics Education I.II: Group 7 222

In Algebra’s 1st Book, Bombelli introduced the terms più di meno (pdm) and meno
di meno (mdm) to represent +i and –i and gave some “basic rules”. Let us consider them
in Bombelli’s original words (p. 169):
“Più via più di meno, fa più di meno. Meno via più di meno, fa meno di meno.
Più via meno di meno, fa meno di meno. Meno via meno di meno, fa più di meno.
Più di meno via più di meno, fa meno Più di meno via men di meno, fa più.
Meno di meno via più di meno, fa più. Meno di meno via men di meno, fa meno.”

Fig. 2: From Bombelli’s Algebra
European Research in Mathematics Education I.II: Group 7 223

Now, let us translate:
"Più" ® +1; "Meno" ® -1; "Più di meno" ® + i; "Meno di meno" ® - i;
"Via" ® · (multiplication); "Fa" ® = .
So we can write:
(+1) · (+i) = +i (–1) · (+i) = –i
(+1) · (–i) = –i (–1) · (–i) = +i
(+i) · (+i) = –1 (+i) · (–i) = +1
(–i) · (+i) = +1 (–i) · (–i) = –1

First of all, we underline the importance of the linguistic aspect: for example, we
have translated the term “Fa” with the symbol “=”; however the modern equality
symbol can be referred to a relation “in two directions”, while the term “Fa” means that
the result of the multiplication in the first member is written in the second member. This
can be confirmed by further research.

Moreover, in Algebra we find (p. 70):
“Più via più fa più. Meno via meno fa più.
Più via meno fa meno. Meno via più fa meno".

We can express these in the following way:
(+1) · (+1) = +1 (–1) · (+1) = –1
(+1) · (–1) = –1 (–1) · (–1) = +1

So we can write the following Cayley table:
x +1 -1 +i -i
+1 +1 -1 +i -i
-1 -1 +1 -i +i
+i +i -i -1 +1
-i -i +i +1 -i

It can be interpreted as the multiplicative group ({+1; –1; + i; – i}; ·) of the fourth
roots of unity, a finite Abelian group (it is well known that it is a cyclic group and it can
be generated either by i either by –i; however Bombelli did not notice explicitly this
property; moreover he did not notice this for its cyclic subgroup generated by –1).
European Research in Mathematics Education I.II: Group 7 224

Of course, in Bombelli’s Algebra we cannot find a modern introduction either of
complex numbers or of the formal notion of a group: Bombelli just indicated some
mathematical objects in order to solve cubic equations. These ideas were not
immediately accepted following the publication of Cardan’s and Bombelli’s works.
Bombelli himself was initially doubtful and wrote in Algebra, p. 169: “I found another
kind of cubic root... and I did not consider it real, until I have found its proof”; let us
underline that Bombelli considered fundamental the geometric proof of algebraic
statements. However the formal introduction of i in Bombelli’s Algebra is important
and modern (Bourbaki, 1960, pp. 91-92). We notice, according to A. Sfard, that
complex numbers were introduced simply as an operational concept:
“Cardan’s prescriptions for solving equations of the third and fourth order,
published in 1545, involved [...] even finding roots of what is today called
negative numbers. Despite the widespread use of these algorithms, however,
mathematicians refused to accept their by-products [...] The symbol [square root
of -1 was] initially considered nothing more than an abbreviation for certain
‘meaningless’ numerical operations. It came to designate a fully fledged
mathematical object only after mathematicians got accustomed to these strange
but useful kinds of computation” (Sfard, 1991, p. 12).

We can consider the idea of group in a similar way: surely it would be incorrect to
ascribe to Bombelli an explicit awareness of the group concept, three centuries before
Galois and Dedekind. However, we can state that he (implicitly) introduced – in action
– one of the most important concepts of mathematics.

2. Educational Problems: The Focus of Our Work

In an important paper, E. Dubinsky, J. Dautermann, U. Leron and R. Zazkis opened “a
discussion concerning the nature of the knowledge about abstract algebra, in particular
group theory, and how an individual may develop an understanding of various topics in
this domain” (Dubinsky, & al., 1994, p. 267).

Let us now consider only the group concept (in the article quoted we can find
interesting considerations of many algebraic notions: the concepts leading up to
“quotient group”, for instance, are: “group”, “subgroup”, “coset”, “coset product” and
European Research in Mathematics Education I.II: Group 7 225

“normality”: Dubinsky & al., 1994, p. 292). The authors write: “An individual’s
knowledge of the concept of group should include an understanding of various
mathematical properties and constructions independent of particular examples, indeed
including groups consisting of undefined elements and a binary operation satisfying the
axioms” (Dubinsky & Al., 1994, p. 268; Leron & Dubinsky, 1995; as regards the
teaching procedure using the computer software ISETL, presented in Dubinsky & Al.,
1994, see for example: Dubinsky & Leron, 1994).

In a recent paper (1996), B. Burn strongly emphasises that the notion of group in
Dubinsky & al. (1994) is introduced by formal definitions. In particular, a group is “a
set with a binary operation satisfying four axioms [...] They espouse a set-theoretic
viewpoint” (Burn, 1996, p. 375). Then Burn notes that “a set-theoretic analysis is a
twentieth-century analysis performed upon the mathematics of earlier centuries as well
as our own” (Burn, 1996, p. 375); so he suggests a pre-axiomatic start to group theory
(Burn, 1996, p. 375; for example, the author quotes: Jordan & Jordan, 1994). So,
according to Burn, the group concept can be introduced before offering axioms (he
points out the importance of geometric symmetries: Burn, 1996, p. 377; Burn, 1985; as
regards fundamental concepts of group theory he quotes: Freudenthal, 1973).

In another recent paper (1997), E. Dubinsky & al. assert that “seeing the general in
particular is one of the most mysterious and difficult learning tasks students have to
perform” (Dubinsky & al., 1997, p. 252; see also Mason & Pimm, 1984); the authors
mention students’ difficulties with permutations and symmetries (Asiala & Al., 1996;
they quote moreover: Breidenbach & Al., 1991; Zazkis & Dubinsky, 1996; as regards
visualisation, see: Zazkis & Al., 1996). So a very important question is the following: is
it possible (and useful) to introduce the group concept by a pre-axiomatic first

In this paper we do not claim to give a full answer to this question: rather we attempt
to contribute to knowledge of the development of students’ understanding of the group
concept, with reference to historical examples. For instance, can the consideration of
the group mentioned above help students in the comprehension of the group concept?
In particular, will consideration of Bombelli’s “basic rules” bring all the properties that
are fundamental to the group concept to pupils’ awareness?

As indicated above, we operate on teaching to improve its quality by thought
European Research in Mathematics Education I.II: Group 7 226

transference; but some reactions, especially those in pupils’ minds, are inferred, they
are plausible rather than certain. We proposed an example in the historical sphere, in
order that students will “learn” in this sphere, but so that the knowledge achieved will
not be confined to the historical sphere: evolution to different spheres is necessary. A
problem that can limit the efficacy of mathematics education as (merely) thought
transference is as follows: if we operate (only) on teaching, are we sure that a correct
evolution will take place in the students?

In what follows we shall examine students’ behaviour in response to a comparative
teaching experiment carried out with two samples of high school students. To the first
sample, we simply quoted Bombelli’s “basic rules”; in the second one we gave the
Cayley table. We wanted to find out if the four properties used in the definition of group
(according to Burn, they are introduced by the terms: “closed”, “associative”,
“identity” and “inverse”: Burn, 1996, p. 372) are acquired by students from these

3. The Group Concept from History to Mathematics Education

The sample comprised the students of three classes from the 3rd Liceo Scientifico
(pupils aged 16-17 years), 68 pupils, and of three classes from the 4th Liceo scientifico
(pupils aged 17-18 years), 71 pupils (total: 139 pupils), in Treviso (Italy). At the time of
the experiment, pupils knew the definition of i (i2 = –1); they did not know the group

We divided (at random) every class into two parts, referred to as A and B; then we
gave the following cards to the students. In the card given to the students in part A (total
68 students) we just quoted Bombelli’s “basic rules”; in the card given to the students of
the part B (total 71 students), we did not quote Bombelli’s “basic rules” but we gave
Cayley table. (See Cards A and B illustrated below.) By the test questions, we wanted to
find out if consideration of a simple historical example (without an axiomatic
set-theoretic viewpoint) is useful in introducing the group concept; moreover, we
wanted to find out the difference between the results obtained by students who were
given a group description by Bombelli‘s “basic rules” and the results obtained by
students who had received the Cayley table.
European Research in Mathematics Education I.II: Group 7 227

As regards the four properties used in the definition of a group, we expected that
closure, associativity and the presence of the unit in the set G = {+1; –1; +i; –i} would
be apparent to many pupils. Closure appears clearly both in Bombelli’s rules and in the
Cayley table; the other properties are quite familiar. The inverse property namely that,
for every xÎG, there is an element x'ÎG such that x·x' = x'·x = 1 can be harder to discern.

Card A. In Algebra, Rafael Bombelli of Bologna (1526-1572) gave the rules:

(+1)·(+1) = +1 (–1)·(+1) = –1 (+1)·(–1) = –1 (–1)·(–1) = +1
(+1)·(+i) = +i (–1)·(+i) = –i (+1)·(–i) = –i (–1)·(–i) = +i

(+i)·(+i) = –1 (+i)·(–i) = +1 (–i)·(+i) = +1 (–i)·(–i) = –1

(we have written the rules by modern symbols).

Consider the set G = {+1; –1; +i; –i}. Are the following statements true or false?
(1) The product of two elements of G is always an element of G.
(2) The multiplication of elements of G is associative.
(3) There is an element e ÎG such that, for every x ÎG, e · x = x · e = x.
(4) For every x ÎG, there is an element x' ÎG such that x · x' = x' · x = e.

Card B. Let us consider the following table:

x +1 -1 +i -i
+1 +1 -1 +i -i
-1 -1 +1 -i +i

+i +i -i -1 +1

-i -i +i +1 -i

Consider the set G = {+1; –1; +i; –i}. Are the following statements true or false?
(1) The product of two elements of G is always an element of G.
(2) The multiplication of elements of G is associative.
(3) There is an element e ÎG such that, for every x ÎG, e · x = x · e = x.
(4) For every x ÎG, there is an element x' ÎG such that x · x' = x' · x = e.
European Research in Mathematics Education I.II: Group 7 228

The time allowed to read the card and answer the questions was 10 minutes. (We
wanted students to examine the problem ‘at a glance’). The students answers were as
follows for the two samples.

Card A true false no answers
(1) 59 87% 2 3% 7 10%
(2) 65 96% 1 1% 2 3%
(3) 48 71% 3 4% 17 25%
(4) 27 40% 21 31% 20 29%

Card B true false no answers
(1) 69 97% 2 3% 0 0%
(2) 70 99% 0 0% 1 1%
(3) 69 97% 0 0% 2 3%
(4) 41 58% 8 11% 22 31%

These results indicate that properties 1, 2, 3 (closure, associativity, presence of the
unit element) are recognised by students. As regards properties 1 and 2 the differences
between sample A and sample B are fairly small, with a greater difference for property
3: it seems that the Cayley table was somewhat more helpful to pupils than Bombelli’s
rules. As regards property 4 (the presence of inverses) the situation is rather different:
only 40% of the students (sample A) and 58% (sample B) accepted the property. Let us
represent some of the results in Tab. I (it is just a qualitative representation: the
differences between percentages in test A and in test B are sometimes slight).

We interviewed individually those pupils (A: 41 students; B: 30 students) who did
not consider statement 4 to be “true”. Almost all these students simply stated that they
did not realize the presence of the inverses just by examining Bombelli’s rules (or the
Cayley table). So consideration of the historical example was successful in causing all
the conjectured reactions and expected effects for only some of the students (let us
notice that it is well known that a multiplicative finite submonoid G of the
multiplicative group C* of non-zero complex numbers is a subgroup: the sufficiency of
the closure test, in this case, is underlined also in: Burn, 1996, p. 373; so, as regards the
previous tests, it is inconsistent to state that properties 1, 2, 3 are true and property 4 is
European Research in Mathematics Education I.II: Group 7 229

not true; however, high school pupils cannot know the mentioned proposition:
Dubinsky & Al., 1997, p. 251).

Comprehension of group axioms percentages

87% 97% 96% 99% 71% 97% 40% 58%
Closure Associativity Unit Inverses

Tab. I

4. History of Mathematics and Epistemology of Learning

The consideration of relevant examples from the history of mathematics can really help
the introduction of important topics. As regards the group concept, however, the
supposed reactions took place completely only for some students.

A pre-axiomatic start to group theory can be useful (see for example: Jordan &
Jordan, 1994), but it is not always enough to assure full learning. As indicated above,
the question remains open. It is possible to object that the mere offering of Bombelli’s
rules is insufficient to achieve a complete learning of the group concept. Let us
emphasise that our research was an exploratory study. For a fuller investigation it
would be necessary to identify clearly sampling criteria and pre-course intuitions (as
underlined in: Burn, 1996, p. 371).
European Research in Mathematics Education I.II: Group 7 230

Let us quote once again Dubinsky & al. (1997):

“A historical view is useful in designing research and instruction with respect
to group theory. History is certainly a part of our methodology, but we are
influenced not only by the record of who proved what and when, but also with
the mechanisms by which mathematical progress was made”
(Dubinsky & al., 1997, p. 252; according to Piaget & Garcia, 1983, there is a
close connection between historical and individual development at the level of
cognitive mechanism; see: Dubinsky & Al., 1997, p. 252).

The main limitation of the notion of mathematics education as thought transference
lies in the uncertainty about real effects (upon the learning) of teachers’ choices. We
make no claims for the teaching of abstract algebra – through the consideration of
historical references or otherwise – as regards the nature and the meaning of
mathematical objects. Here several problems are opened (Godino & Batanero, 1998),
involving several fundamental philosophical questions. However it is important and
necessary to control the educational research process by experimental verification: this
can profoundly affect the delineation of the research and give it an important, particular
epistemological status.

5. References
Asiala, M.; Brown, A.; DeVries, D.; Dubinsky, E.; Mathews, D.; & Thomas, K. (1996): A framework
for research and curriculum development in undergraduate Mathematics Education, Research
in Collegiate Mathematics Education, 2, 1-32.
Bourbaki, N. (1960): Eléments d’histoire des mathematiques, Paris: Hermann.
Breidenbach, D.; Dubinsky, E.; Hawks, J. & Nichols, D. (1992): Development of the process
conception of function, Educational Studies in Mathematics, 23, 247-285.
Burn, B. (1996): What are the fundamental concepts of group theory?, Educational Studies in
Mathematics, 31, 371 -377.
Burn, R.P. (1985): Groups: a Path to Geometry, Cambridge.
Dubinsky, E.; Dautermann, J.; Leron, U. & Zazkis, R. (1994): On learning fundamental concept of
group theory, Educational Studies in Mathematics, 27, 3, 267-305.
Dubinsky, E. & Leron, U. (1994): Learning Abstract Algebra with ISETL, New York: Springer.
European Research in Mathematics Education I.II: Group 7 231

Dubinsky, E.; Dautermann, J.; Leron, U. & Zazkis, R. (1997): A reaction to Burn’s “What are the
fundamental concepts of group theory?”, Educational Studies in Mathematics, 34, 249-253.
Freudenthal, H. (1973): What groups mean in mathematics and what they should mean in
mathematical education, Howson, A.G., Developments in Mathematical Education, ICME-2,
101-114. Cambridge: Cambridge University Press.
Godino, J.D. & Batanero, C. (1998): Clarifying the meaning of mathematical objects as a priority area
for research in Mathematics Education, Sierpinska, A. & Kilpatrick, J., Mathematics Education
as a Research Domain: a Search for Identity, 177-195. Dodrecht: Kluwer.
Jahnke, H.N. (1991): Mathematik historisch verstehen, Mathematik Lehren 47, 6-12.
Jahnke, H.N. (1995): Historische Reflexion im Unterricht, Mathematica Didactica 18, 2, 30-58.
Jahnke, H.N. (1996): Mathematikgeschichte far Lehrer, Gründe und Beispiele, Mathematische
Semesterberichte 43, 1, 21-46.
Jordan, C. & Jordan, D. (1994): Groups, London: Arnold.
Leron, U. & Dubinsky, E. (1995): An abstract algebra story, American Mathematical Monthly, 102, 3,
Mason, J. & Pimm, D. (1984): Generic examples: seeing the general in the particular, Educational
Studies in Mathematics, 15, 277-289.
Piaget, J. & Garcia, R. (1983): Psychogenèse et histoire des sciences, Paris: Flammarion.
Sfard, A. (1991): On the dual nature of mathematical conceptions: reflections on processes and
objects as different sides of the same coins, Educational Studies in Mathematics, 22, 1-36.
Zazkis, R. & Dubinsky, E. (1996): Dihedral groups: a tale of two interpretations, Research in
Collegiate Mathematics Education, 2, 61-82.
Zazkis, R.; Dubinsky, E. & Dautermann, J. (1996): Coordinating visual and analytic strategies: a
study of students’ understanding of the group D4, Journal for research in Mathematics
Education, 27, 4, 435-457.
European Research in Mathematics Education I.II: Group 7 232

Juan Diaz Godino, Carmen Batanero
University of Granada, Facultad de Educación, Campus de Cartuja, 18071 Granada, Spain

Abstract: In this report we argue that the notion of meaning, adapted to the specific
nature of mathematics communication, may serve to identify analysis units for
mathematical teaching and learning processes. We present a theory of meaning for
mathematical objects, based on the notion of semiotic function, where we distinguish
several kinds of meanings: notational, extensional, intensional, elementary, systemic,
personal and institutional. Finally, we exemplify the theoretical model by analysing
some semiotic processes which take place in the study of numbers.
Keywords: ontology, semiotics, mathematics education.

1. Introduction

According to Vygotski (1934), the unit for analysing psychic activity - which reflects
the union of thought and language - is the meaning of the word. This meaning -
conceived as the generalisation or concept to which that word refers - englobes the
properties of the whole, for which its study is considered, in its simpler and primary
form. We think that the search of analysis units for mathematical teaching and learning
processes should also be focussed on the meaning of the objects involved.

However, the notion of meaning should be interpreted and adapted to the nature of
mathematical knowledge and to the cognitive and sociocultural processes involved in
its genesis, development and communication. We agree with Rotman (1988) in that it is
possible and desirable to develop a specific semiotic of mathematics, which would take
into account the dialectics between mathematical sign systems, mathematical ideas and
the phenomena, for the understanding of which they are built, within didactic systems.
European Research in Mathematics Education I.II: Group 7 233

To reach this aim, we consider it necessary to elaborate a notion of meaning specifically
adapted to didactics, by interpreting and adapting existing semiotical and
epistemological theories. This need was forecast by Brousseau (1986 p. 39), who
wondered whether there is or there should be a notion of meaning, unknown in
linguistics, psychology, and mathematics, though especially appropriate for didactics.

In this research work, we present some elements of a semiotic model, specific to
Didactic of Mathematics, starting out from the notion of semiotic function proposed by
Eco (1979), and classifying mathematics entities into three types: extensional,
notational and intensional ones. Based on the different nature of these types of
mathematical entities and the contextual factors conditioning mathematical activity, we
identify meaning categories to describe and explain the interpretation and
communication processes taking place in the heart of didactic systems.

The notion of meaning -conceived as the content of semiotic functions and
applicable to mathematical terms and expressions, as well as to conceptual objects and
problem situations - allows us to identify analysis units for mathematics teaching and
learning processes. We finally describe some of these units, applying them to analyse
examples of semiotic processes at the different stages of the study of whole numbers.

2. Meaning as the Content of Semiotic Functions

We use the term ‘meaning’ according to the theory of semiotic functions described by
Eco (1979), and consider the pragmatic context as a conditioning factor of such
semiotic functions. Here, this context includes the set of factors sustaining and
determining mathematical activity, and therefore, the form, appropriateness and
meaning of the objects involved. According to Eco, “there is a semiotic function when
an expression and a content are in correlation” (Eco 1979 p.83). Such a correlation is
conventionally established, though this does not imply arbitrariness, but it is
coextensive to a cultural link. There may be functives of any nature and size. The
original object in the correspondence is the signifiant (plane of expression), the image
object is the meaning (plane of content), that is, what it is represented, what it is meant,
and what is referred to by a speaker.
European Research in Mathematics Education I.II: Group 7 234

In a previous article (Godino and Batanero 1998) we analysed the emergence of
mathematical objects from the meaningful practices carried out by persons or
institutions when solving specific problem fields. A meaningful practice is defined as
“a manifestation (linguistic or not) carried out by somebody to solve mathematical
problems, to communicate the solution to other people, to validate and generalise that
solution to other contexts and problems ”, and the meaning of a mathematical object is
identified as the system of practices linked to the field of problems from which the
object emerges at a given time.

Since semiotic functions are established by a person in a given context with a
communicative or operative intention, for us these functions can be considered as
meaningful practices and reciprocally, behind each meaningful practice we could
identify a semiotic function or a lattice of semiotic functions.

Meaningful practices, conceived as intentional actions mediated by signs, might be
the basic units for analysing cognitive processes in mathematics education.

3. Notational, Extensional and Intensional Meanings
We can establish semiotic functions between three primitive types of mathematics
• Extensional entities are the problems, phenomena, applications, tasks, i.e., the
situations which induce mathematical activities.
• Notational entities, that is, all types of ostensive representations used in
mathematical activities (terms, expressions, symbols, graphs, tables, etc.)
• Intensional entities: mathematical ideas, generalisations, abstractions (concepts,
propositions, procedures, theories).

In Godino and Recio (1998), we analyse to some extent the nature of these entities
interpreting and adopting ideas by Freudenthal, Vergnaud and Dörfler. Notational
entities play the role of ostensive and essential support that makes mathematical work
possible, because generalisations and situation problems are given by notational
systems, which describe their characteristic properties. Abstractions are not directly
observable and problem situations are frequently used to provide abstract mathematical
European Research in Mathematics Education I.II: Group 7 235

objects with a context. Abstractions and situations are neither inseparable from the
notations (ostensive objects) which embody them, nor identifiable with them, that is,
we consider that mathematics cannot be simplified to the language which expresses it.

The three types of primary entities considered (extensional, intensional and
notational) could perform both roles of expression or content in semiotic functions.
There are, therefore, nine different types of such functions, some of which may clearly
be interpreted as specific cognitive processes (generalisation, symbolisation, etc.).

In this paper, we classify and characterise these functions as regards to the content
(meaning) involved, so that the nine types are reduced to the following three:

(1) Notational meaning: Let us call a semiotic function notational when the final
object (its content), is a notation, that is, an ostensive instrument. This type of function
is the characteristic use of signs to name world objects and states, to indicate real things,
to say that there is something and that this is built in a given manner. The following
examples demonstrate this type of meaning:
• When a particular collection of five things are represented by the numeral 5.
• The symbol Pn (or n!) represents the product n(n-1)(n-2)...1

(2) Extensional meaning: A semiotic function is extensional when the final object is
a situation problem, as in the following examples:
• The simulation of phenomena (i.e., it is possible to represent a variety of
probabilistic problems with urn models).

(3) Intensional meaning: A semiotic function is intensional when its content is a
generalisation, as in the following examples:
• In expressions such as, “Let m be the mathematical expectation of a random
variable ”, or “ Let f (x) be a continuous function”. The notations m, f(x), or the
expressions ‘mathematical expectation’, ‘random variable ‘ and ‘continuous
function’, refer to mathematics generalisations.

Furthermore, all intensional and extensional functions imply an associated
notational function, since abstractions as well as problem situations are textually fixed.
European Research in Mathematics Education I.II: Group 7 236

4. Elementary and Systemic Meanings

An elementary meaning is produced when a semiotic act (interpretation /
understanding) relates an expression to a specific content within some specific
space-temporal circumstances: It is the content that the emitter of an expression refers
to, or the content that the receiver interprets. In other words, what one means, or what
the other understands. Examples of this use of the word ‘meaning’ are the deictic signs,
rigid designations (Eco 1990) where the content is indicated by gestures, indications or
proper names. The content of the semiotic function is a precise object, which may be
determined without ambiguity in the spatial-temporal circumstances fixed.

The semiotic processes involved in building mathematical concepts, establishing
and validating mathematics propositions, and, as a rule, in problem solving processes,
yield systemic meanings. In this case, the semiotic function establishes the
correspondence between a mathematical object and the system of practices which
originates such an object (Godino and Batanero 1998). The structural elements of this
systemic meaning would be the problem situations (extensional elements), the
definitions and statements of characteristic properties (intensional elements) and the
notations or mathematical registers (notational elements). These three types of
primitive entities provide a classification of practices constituting mathematical

5. Personal and Institutional Meanings

The theoretical nature of systemic meanings and encyclopaedias tries to explain the
complexity of semiotic acts and processes, but they are not fully describable. Practice
systems differ substantially according to the institutional and personal contexts where
problems are solved. These contexts determine the types of cultural instruments
available and the interpretations shared, and therefore the types of practices involved.

“Even when, from a general semiotic viewpoint, the encyclopaedia could be
conceived as global competence, from a sociosemiotic view is interesting to
determine the various degrees of possession of the encyclopaedia, or rather the
European Research in Mathematics Education I.II: Group 7 237

partial encyclopaedias (within a group, sect, class, ethnic groups, etc” (Eco
1990, p. 134).

Due to these characteristics of the systemic meanings, we consider it necessary to
distinguish between institutional meanings and personal meanings, depending on
whether practices are socially shared, or just idiosyncratic actions or manifestations of
an individual. In the second case, when the subject tries to solve certain classes of
problems, he builds a personal meaning of mathematical objects. When this subject
enters into a given institution (for example, the school) he/she might acquire practices
very different from those admitted for some objects within the institution.

A matching process between personal and institutional meanings is gradually
produced. The subject has to appropriate the practice systems shared in the institution.
But the institution should also adapt itself to the cognitive possibilities and interests of
its potential members.

The types of institutions interested by a specific class of mathematical problems
might be conceived as communities of interpreters sharing some specific cultural
instruments and constitute a first factor for conditioning the systemic meanings of
mathematical objects.

6. Semiotic Acts and Processes in the Study of Numbers

In this section, and to give examples of the theoretical concepts described, we apply the
semiotic model outlined to the analysis of some semiotic acts and processes involved in
the study of whole numbers.

6.1 Elementary Meanings: The First Encounter With Numbers

The first encounter with numbers for most children, is produced at pre-school age,
when their parents teach them the series of words ‘one’, ‘two’, ‘three’, etc. to count
small collections of objects: hand fingers, balls, sweets, etc. Afterwards, they will find
European Research in Mathematics Education I.II: Group 7 238

school tasks close to those reproduced in Fig.1, which have been taken from a book
for1st year primary teaching.

The text is intended to make the child recognise and write the numerals ‘1’, ‘2’, ‘3’,
..., at the same time as different collections of objects represented are assigned to the
corresponding numerical symbol. From the drawing of a head, a sun, a cat an arrow
points at to the symbol 1. As an exercise, drawing 1 beside a flower and a moon is
implicitly requested. A similar method is used for teaching the number 2, its form,
writing, and use.

In the tasks proposed, we can identify the three classes of objects and semiotic
functions which characterise mathematical activity, according to our
semiotic-anthropological model: Notations (ostensive instruments), extensions, and
generalisations (or abstractions).

Fig. 1: Learning the numbers 1 and 2

In fact, the concrete object drawings (head, sun, cat, flowers, eyes, etc.) are iconic
representations of such objects; the meaning of the icons is the corresponding concrete
object (extensional meaning); the schemes implicitly suggest answering the question,
European Research in Mathematics Education I.II: Group 7 239

“How many objects are there?”, so that the reference context is not just made up by
mere concrete objects, but rather the problem situation of computing the cardinal. It is
important to observe that, in the task proposed by the book, the physical objects are not
in fact present. Therefore, the immediate reference context to which numerical symbols
refer is a world of ostensive representation (textual). This fact implies additional
semiotic complexity, and hence, interpretative effort by the child. In this series of tasks,
we also identify two operative invariants (generalisations):

• the same symbol,’1’, and the number-word ‘one’ are associated to various
drawings of unitary collections;
• the symbol, ‘2’, and the number-word ‘two’ are associated to different pairs of
objects. The mental objects (or better, the logical entities), one and two, are
implicitly evoked as from the first teaching levels. The tasks aim, in
psychological terms, is the progressive construction of the objects, number one,
two, etc. in the child’s mind. In anthropological terms it is the child’s acquisition
of the habit of naming any collections by using the series of number-words, and
the series of numerical-symbols (numerals).

We also identify the following semiotic functions (acts and interpretation

I1: The drawing of concrete objects is implicitly related to the concrete objects
(extensional meaning) they represent.
I2: The object collections are interrelated with the numerical symbols, ‘1’, ‘2’
(notational meaning).
I3: The number-words ‘one’, ‘two’, are associated (implicitly) with the numerical
symbols, ‘1’, ‘2’ (notational type).
I4: Each icon collection, and its associated numerical symbol, is implicitly
interrelated with the corresponding mathematical concepts (one, two, three, ...)
(intensional semiotic function).

The interpretation of these semiotic functions requires specific codes and
conventions which should be known and interpreted by the receiver of the message (the
child), to successfully complete the tasks and to progressively acquire the notion of
European Research in Mathematics Education I.II: Group 7 240

number. We point out that the spaces left between the different icons play a
fundamental role, since they inform which objects should be counted in each case.

The writing guides for learning to draw the numerals ‘1’ and ‘2’ are also full of
symbolisms, which they are difficult to decode without the teacher’s assistance. They
graphically present rules such as: “Do this drawing in the way I show you and repeat it
several times”.

This type of rules poses the subject with a problem situation (or simply, a routine
task), and involves an extensional meaning, according to our theoretical model.

Our analysis shows the multiplicity of codes for whose recognition the children will
require a systematical teacher’s assistance. This supports Solomon’ s thesis (1989 p.
160), that “knowing number should be reconceptualized as involving entering into the
social practices of number use”, and not as an issue of individual construction of the
necessary and sufficient logical structures for understanding numerical concepts.

6.2 Systemic Meanings of Numbers

In the previous section we have shown examples of notational, extensional and
intensional elementary meanings involved in the study of numbers at elementary
school. The organised set of these elementary meanings would correspond to what we
call a number systemic meaning, which would be personal or institutional depending
on whether we take a particular subject (a child) or an institution (community of
interpretants) as a reference.

The systemic meaning of numbers within a given teaching level (school institution)
is determined by curricular documents, school textbooks, and by the teachers’ own
preparation of their lessons on a mathematical topic.

We can observe that the meaning of numbers in curricular documents, is described
in an encyclopaedic or systemic form, since it refers to a complex of situational,
intensional and notational elements. Numerical competence will be achieved through
European Research in Mathematics Education I.II: Group 7 241

carrying out an organised practices system of progressive complexity throughout a
prolonged period of time, which is extended beyond primary teaching.

The institutional systemic meaning of numbers described in the curricular
documents, will later be interpreted by the textbook authors and by the teachers
themselves when designing their didactic interventions, to select and fulfil the practices
they consider most appropriate to their institutional circumstances. These practices will
finally be carried out by the pupils, and will determine the personal meanings that these
pupils progressively build.

At primary school, numbers are some “special symbols”, 1, 2, 3, ... associated to
collections of objects, to count, order, and name them. Children may also carry out
activities with concrete materials (rods, toothpicks, multibase blocks, abaci) which
constitute more primitive numbering systems than the place-value decimal numbering
system, privileged by mathematical culture due to its efficiency. It is not rare, therefore,
that if we ask a child, “What are numbers?”, he/she will answer, at best:
“They are symbols, 0, 1, 2, 3,. ...., invented by man to count and compare quantities.
These symbols form the set of numbers”.

At elementary school, numbers are neither ‘the cardinal of finite sets’, nor ‘the
common property to all finite sets mutually coordinable. Few people (children, adults,
even teachers) would provide such a description of numbers; however, they handle
numbers, know to use them effectively for counting and ordering. For these people,
numbers have a different meaning from that shared by professional mathematicians.

Even for professional mathematicians the descriptions of numbers may vary
substantially. In Cantorian mathematics, whole numbers “are the elements of the
quotient set determined on the set of finite sets by the relationship of equivalence of
coordinability between sets”. However, for Peano’s mathematics, a totally ordering set
will be called whole numbers if it fulfils the following conditions:

• Any successor of an element of N belongs to N.
• Two different elements of N cannot have the same successor.
• There is an element (0) that it is not a successor of any other element in N.
European Research in Mathematics Education I.II: Group 7 242

• All subsets of N that contain 0 and contain the successor of each one of their
elements coincide with N.

Therefore, there is no single definition of whole number set, but rather various,
adapted to problem situations, intentions and semiotic tools available in each particular
circumstance. Each definition we may make for whole numbers emerges from a
specific practices system, hence, it involves a class of problem situations and specific
notational systems. In principle, each phenomenological numerical context (sequence,
counting, cardinal, ordinal, measure, label, number writing, computation) can produce
an idiosyncratic meaning (or sense) for numbers. Institutional contexts also share
idiosyncratic practice systems and, consequently, they determine differentiated

7. Conclusions and Implications

The idea guiding our work is the conviction that the notion of meaning, in spite of its
extraordinary complexity, may still play an essential role to as a basis for research into
the didactic of mathematics. We think that an anthropological approach to this
discipline, as Chevallard (1992) proposes, complemented with specific attention to
semiotic processes, may help us to overcome a certain transparency illusion about
mathematics teaching and learning processes, showing us the multiplicity of codes
involved and the diversity of contextual conditioning factors. The construct of systemic
meaning postulates the complexity of mathematical knowledge by recognising its
diachronic and evolutionary nature. This makes us aware of the relevance of
semiotic-anthropological analysis of problem fields associated to each knowledge,
their structure variables, and the notational systems used, since knowledge emerges
from people’s actions when faced with problem situations, as mediated by the semiotic
tools available.

Hence, it may be useful in curricular design, development and evaluation as a
macro-didactic unit of analysis, guiding the search and selection of representative
samples of practices characterising mathematical competence.
European Research in Mathematics Education I.II: Group 7 243

On the other hand, and taking into account the complex nature of the meaning of
mathematical objects, we should often focus attention on specific interpretative
processes and on the inherent difficulties of the same when analysing student and
teacher classroom performances. The construct of elementary meaning and the
description of its various types permits us to focus attention on the implicit codes which
condition acts and processes of understanding in mathematics education. This will be
useful for identifying critical points, conditioning factors of semiotic acts and processes
in mathematical activity and anticipating didactical actions.

The meaning of a mathematical object has a theoretical nature and cannot be totally
and unitarily described. Practices carried out to solve mathematical problems differ
substantially according to institutional and personal contexts. For this reason, we
introduce institutional meanings to distinguish between these different points of view
and uses on the same mathematical object. These practices are interpreted in this article
as semiotic functions or sequences of semiotic functions, analysing their types, and
taking into account the nature of their content (extensional, intensional and notational).

Personal meanings are built by the individual subject - what he/she learns, his/her
personal relation to the object - and do not just depend on cognitive factors, but rather
on the semiotic-anthropological complex in which this relation is developed, that is, on
the element of meaning and the dialectic relationships between them as they are

The theoretical model outlined intends to facilitate the study of the relationship
between personal and institutional meanings of mathematical objects. It also implies a
strong support for conceiving mathematics and its teaching and learning as a social
practice. Children’s learning difficulties, errors or failures, are explained by their
different interpretations of each situation with respect to what the teacher intended, or
simply by their lack of familiarity with the situation. “Not having entered into the social
practices of a particular situation, subjects are lost about how to act, though they make
the best of it”. (Solomon 1989, p. 162).
European Research in Mathematics Education I.II: Group 7 244

8. References
Brousseau, G. (1986): Fondements et méthodes de la didactiques des mathématiques. Recherches en
Didactique des Mathématiques, Vol. 7, nº 2: 33-115.
Chevallard, Y. (1992): Concepts fondamentaux de la didactique: perspectives apportées par une
approche anthropologique. Recherches en Didactique des Mathématiques, Vol. 12, nº 1:
Eco, U. (1979): Tratado de semiótica general. Barcelona: Lumen.
Eco, U. (1984): Semiótica y filosofía del lenguaje. Madrid: Lumen, 1990.
Godino, J. D. & Recio, A. M. (1998): A semiotic model for analysing the relationships between
thought, language and context in mathematics education. Research Report Proposal. 22nd PME
Conference (South Africa).
Godino, J. D. & Batanero, C. (1998): Clarifying the meaning of mathematical objects as a priority area
of research in mathematics education. In A. Sierpinska & J. Kilpatrick (Eds.), Mathematics
Education as a research domain: A search for identity. pp. 177-195. Dordrecht: Kluwer, A. P.
Rotman, B. (1988): Towards a semiotics of mathematics. Semiotica, 72 -1/2: 1-35.
Solomon, Y. (1989): The practice of mathematics. London: Routledge.
Vygotski, L. S. (1934): Pensamiento y lenguaje. [Obras escogidas II, pp. 9-287]. Madrid: Visor, 1993.

This research has been supported by the DGES (MEC, Madrid), Project PB96-1411.
European Research in Mathematics Education I.II: Group 7 245

José Luis González Marí
Department of Didactic of Mathematics. University of Málaga, Campus Teatinos s/nº, 29071 Málaga, Spain

Abstract: The methods used for the research in Mathematical Sciences Education are
common for Psychology, Pedagogy and other related fields. But, to be honest, many of
these approaches lead to too punctual results, not very important and with few
possibilities to modify substantially the educational practice. The reason could lies in
the inadequacy of such methods to cover the complexity of the field, in which many
factors are operating in a non isolated way and interconnected to each other through
relationships which need to be identified and analysed previously in a global
framework under the mathematical knowledge as a common factor. To carry out this
previous task, about which we must decide later the usage of the most suitable methods,
we have been using a non empirical integrating research procedure, called Didactical
Analysis, built up in the conjunction of meta-analysis and qualitative approaches. This
paper exposes the principles, the conceptual framework and the techniques that make
up this still being studied methodology.
Keywords: methodology, meta-analysis, epistemology of mathematics.

1. Introduction

The methodology deals, in a general way, with the way we get knowledge about the
world (Denzin; Lincoln, 1994). In this case, it deals with the world of the phenomena in
the educational processes of teaching and learning of Mathematics. The study of these
phenomena can be raised by following a general process which Romberg (1992, p. 51)
summarizes in ten steps represented in bold line’s squares in figure 1. From these ten
main activities, the author attaches special importance to the first four ones, while the
last six ones (from 5 to 10) have to do with the operating or technical part of the process.
But, as long as there are more than twenty recognized procedures for developing this
European Research in Mathematics Education I.II: Group 7 246

second part, we find that the activities expected to be the most important ones in the
process seem to be supported, solely and without any added specifications, by the
intuition and researcher’s skills, by the background of the problem to be researched
(Bishop, 1992, pp. 712-714) and by the previous knowledge coming from the scholarly
community. These considerations would be enough to guarantee the quality and
relevance of the results if the phenomena were not so complex as they really are, if the
main factors of that complexity took part of the research in terms of relationships and,
finally, if scholars bore in mind a kind of integration of the different perspectives and
traditions beyond the mere interdisciplinary approach. But these conditions either are
not carried out satisfactorily or there are doubts that it would be like that, which may
lead to the so called “feet of clay” research, that is to say, faultless from the operative or
technical point of view (activities 5 to 10) but faulty regarding their initial fundamentals
and assumptions (activities 1 to 4).

To try to improve this situation in those researches involving a specific
mathematical knowledge, we propose to introduce a mechanism of systematic control
of the research process fundamentals just between activities 3 and 4 of Romberg´s
scheme. It consists in three activities (3.1, 3.2 and 3.3, figure 1) involving a method that
we have recently built and used with acceptable results (González 1995, 1998; Ortiz
1997) looking for an adequate response to the complexity and specificity of
Educational phenomena in Mathematics.

The procedure is based on the general principles of meta-analysis (Glass & McGaw
& Smith 1981), the multivocal revision (Ogawa & Malen 1991) and what is known as
conceptual analysis (Scriven 1988). For its name we have chosen the term “didactical
analysis”, used in other sense to describe “… the analysis of the mathematics contents,
which is carried out at the service of the organisation of its teaching in the educational
systems…”(Puig, 1997, p. 61). In this case we refer to a systematic, genuine and
integrating process which gives an own personality to the research in its first steps.

The modification proposed is justified by the following arguments, which will be
extensively developed in the next sections: a) the complexity of phenomena, in which
many factors are interacting and for which analysis many perspectives and procedures
are needed; b) the partial specificity of the field of study, founded in the involvement of
European Research in Mathematics Education I.II: Group 7 247

mathematical knowledge; c) the lack of the interdisciplinary approach and the need of
an integration of knowledge and approaches in order to achieve a greater effectiveness.

1. Identify a phenomenon Research activities and how they
of interest are related according to
Romberg, T., 1992, p. 51
2. Build a preliminary Our additional notes and activities
model to include the multiple views and
to previously study its implicit
didactical relationships by means
3. Relate to others’ of the Didactical Analysis.
referred to specific
ideas Procedure to face the complexity
case and community
of Math. Education, to gain
objectivity and to integrate the
3.1. Gather information several research traditions
from related fields (*) (Bishop, 1992).
(*) About the different faces and components of
the phenomenon as seen from the several related
3.2. Carry Didactical
fields: Epistemology, Phenomenology, Cognition,
Analysis out (**) Curriculum, Teaching, etc.
(**) A non empirical cualitative synthesis
method (González, 1995; Ortiz, 1997).
3.3. Rebuild the model.
Priorities and decisions

4. Questions or 5. Select research
conjectures strategy

6. Select research 7. Gather
procedure evidences

8. Interpret 9. Report
evidences results

10. Anticipate
actions of others

Fig. 1: Research activities and how they would be related
according to our experience
European Research in Mathematics Education I.II: Group 7 248

2. Mathematics Education: A Field of Relationships vs. Related

In Mathematical Sciences Education, a series of areas, which in the educational
practice interact and operate together, can be identified and theoretically separated. Of
all these areas we can emphasize, firstly, the one that has to do with cognitive aspects,
covering, among others, the characteristics and evolution of learns, errors, difficulties
and representations and the acquisition of automatism and skills.

On the other hand, we can find a field focussed on teaching, in which specific
aspects such as the following: nature, relationships, structure and organisation of the
school curriculum (objectives and other organizers (Rico, 1997, pp. 39-59)) witnessing
complex factors and conditions (sociocultural, economical, etc.), educational policies,
curricular projects and teachers training, can be identified.

Thirdly, we can distinguish a parcel, which is more connected to the real teaching
and learning processes, in which some interactions among different factors from the
two previous fields are taking part at different levels (design, planning and
implementation) (Coriat 1997, pp. 156-157), that is to say: methods for improving
learning; resources and materials and curricular adaptations, for instance.

The separation between the above mentioned three fields, which can mainly be
observed in the preponderance that researches give to the psychological or pedagogical
aspects, seems to be an inadequate approach. The mentioned fields are interconnected
to each other and specially related to the Psychology of Mathematics Education
(Fischbein, 1990, pp. 6-12) in a first level focussed on the educational finalities and on
the general characteristics of mathematical knowledge. Secondly, in special when a
specific mathematical topic is taking part, that first level has a tight dependency from
other basic elements, as it is the case of Mathematics, its Epistemology and its History
or of the Phenomenology of mathematical knowledge; a second level of dependency,
which is focussed on both finalities and mathematical contents and based on the
following general principles (made up from the considerations of Tymoczko (1986),
Davis & Hersh (1988) and Puig (1997)):
European Research in Mathematics Education I.II: Group 7 249

a) Mathematical knowledge is perfectible, partial and incomplete, being exposed to
failures and having to do with ideas or conceptual objects to which humans accede
through the discovery or the non-arbitrary invention or creation. These objects are
independent of its symbolisation, having a conventional existence and sharing two
different areas: the individual conceptual and the supra individual, cultural or
collective as a part of the shared conscience.
b) The phenomena that organize the mathematical concepts are the objects, its
properties, the actions on them and the properties of such actions, belonging, all of
them, to an unique world in expansion containing the results from human cognition
and, in particular, the results from the mathematical activity (Puig, 1997, p. 67).
c) The creation / discovery of mathematical knowledge is conditioned by the common
factors to all individuals and cultures that make it possible: the common
characteristics of human mind (physiological, for instance), of the environment
(physical, social, cultural, for instance) and of the interaction between them.

From the principles and its relationships we draw the following consequences:
1. The involvement of the three factors (mind, environment and interaction) took part
in all the interpretations about nature and the way of production of mathematical
knowledge, so that the epistemological analysis must take its characteristics into
2. The analysis of mathematical knowledge from an educational perspective must
include the epistemological, cognitive and phenomenological analysis, which
would be related to the sociocultural aspects as well as to teaching and curricular
issues as specific and terminal subjects in Mathematical Sciences Education; five
major fields which must be involved in the general research framework.
3. The epistemological and phenomenological analysis regarding with educational
research must follow a markedly didactical line. The interest must be focussed in
obtaining valuable information for teaching and learning, what means thinking
about the student, his needs and capacities, about the classroom, its activities and
didactical methods. The information gathered through this approach shows the links
among the different parts of the two levels mentioned above under a unique
reference: the individual and collective mathematical thought, its evolution, its
relationships with other kinds of thought and its education.
European Research in Mathematics Education I.II: Group 7 250

4. This way, the relationships between the Epistemology of Mathematics and the
Psychology of Education are in a privileged position, and when focussing all the
attention in the processes of creating knowledge it makes sense as a part closely
related to mathematical knowledge and to curricular decisions. Likewise, the
pedagogical aspect shows a close dependence on the previous factors, to which we
should add some other considerations either social, political or cultural that
complete a specific and global world view in which multiple relations are needed of
a previous integration to carry out the particular studies in the different fields and

But, such integration must not be finished in a sum of data obtained from different
approaches (interdisciplinary conception). On the contrary, it requires a complex
elaboration by following a specific research methodology (Didactical analysis) that it
must be quite different from the corresponding to the technical part of the process
(activities 4 to 10, figure 1). Let´s go to treat next on this argument.

3. Insufficiency of the Interdisciplinary Approach and

The insufficiency of the interdisciplinary approach regarding Mathematical Sciences
Education research, which still has nowadays a wide acceptation by the community
although its results can be improved on (Kilpatrick, 1994, pp. 78-79), is based on the
relationships among the five major fields and on the nature of the phenomena.

Firstly, the analysis of the relationships may gives new and genuine nuances to the
isolated information; this nuances gives cohesion to the study and they provides the
integration character we are postulating. Nevertheless, the needs of integration, even
importants, not only are justified because of the characteristics of dependencies
relations; they can also arise because of the existence of a stagnation of the researches
or because of the advanced situation of studies, as it is the case of the concept of
function (Harel & Dubinsky, 1992; Romberg, Fennema & Carpenter, 1993) or of the
rational numbers (Carpenter, Fennema & Romberg, 1993).
European Research in Mathematics Education I.II: Group 7 251

As regards to the nature of the phenomena, we are maintaining the existence of a
specific part in the Mathematical Science Education field, to see which it is only needed
to analyse the answers to the following questions: Are there important differences
between the knowledge of the phenomena in Mathematical Sciences Education when
these are analysed from the particular approach of the Didactics of Mathematics and
when it is carried out in the context of the general educational phenomena?; Where do
such differences lie in?; Is Mathematical Sciences Education a part, in the inclusive
sense, of the general field of Education?; If there are any differences, is it correct to use
the methods which are usually applied for the non-specific educational research?; Are
such methods enough?; Are they priority ones or of preferential application to any other

It is not difficult to come to the conclusion that what is characterizing the
Mathematical Sciences Education is not the interdisciplinary aspect just like that, but a
specific and deeper way of studying the phenomena regarding the teaching and learning
processes. This particular point of view can be summarized, in one hand, in the
involvement of some specific basic components (the sociocultural as general one and
the four remainders as specific and central ones), which have an essential role in the
curricular studies (Rico 1997) and among which we emphasize the one that has to do
with the epistemological and phenomenological considerations about the
mathematical knowledge in a global framework under didactical purposes. In the other
hand, in the delimitation and analysis of a relationships’ network among the four
central components (See figure 2) (Gonz·lez et al. (1994); Gonz·lez (1995, 1998)).
These two aspects do not intervene in the interdisciplinary approach, which is usually
limited to a simple collection of data coming from different approaches.

To sum up: The phenomena regarding the teaching and learning of mathematics
show general aspects, which are part of the interest of other disciplines, and specific
connotations that introduce differences in the way of approaching the same problems
from other fields of knowledge.

Such specificity lies in the important involvement of epistemology and
phenomenology of mathematical knowledge in a global framework under educational
purposes as well as in its relationships to other fields (pointed out by Vergnaud (1990,
p. 22-23)). When taking this into account we can observe an inversion in the
European Research in Mathematics Education I.II: Group 7 252

assumptions considering Didactics of Mathematics as a specialised branch of General
Didactics or of Psychology of Education (Fischbein 1990, pp.6-12).

It is necessary to include a double point of view in the usual research process: a
genuine, specific approach, to support and organize the field (activities 3.1, 3.2 and 3.3,
figure 1) through a also specific procedure, and a later common interdisciplinary
approach to inquiry in the punctual aspects deduced from the previous study (activities
5, 6 and 7, figure 1). Although suitable for particular purposes, the last approach is
neither enough nor priority one, but it must be dependent on the results from the
previous analysis.

projection of information under
Epistemology didactical purposes

Curriculum Relationships areas:
a: Epistemology and
Phenomenology phecu and Teaching Phenomenology
b: Epistemology and
ac c Cognition
c: Epistemology and
Curriculum and Teaching
ab, ac, bc: three components
abc: the four components
a abc bc arrows: from low to high
ab information level

fco cocu

Fig. 2: Relationships’ central network

4. Didactical Analysis of a Mathematical Knowledge

In the studies known as secondary or of synthesis researches two different
methodologies are being used: the traditional integrating revision and the quantitative
revision, also called meta-analysis (Fernández, 1995, p. 165)). Recently, due to the
need aroused in many qualitative researches of summarize and integrate a great number
of studies, a modality of synthesis, called multivocal revision (Ogawa & Malen, 1991)
European Research in Mathematics Education I.II: Group 7 253

has arisen. It is a procedure of qualitative synthesis “…aimed at making inquiries on a
complex phenomenon of interest in which events can not be manipulated and of which
there are many sources of essentially qualitative data, being confident that we can
obtain a detailed portrait of the phenomenon under study”. (Fernández 1995, p. 175).
The multivocal revision is based on the following criteria, which are similar to the ones
suggested for the case studies (p. 176):

1) A clear definition of the research topic by consulting many sources, by keeping
evidence’s chaines among the records and the inferences drawn and by formally
including the informer’s reactions to the established conceptual definition.
2) Assess the relative and individual strength of each piece of information by using
some of the following criteria: position and certainty of the source (external
validity); clearness, detail, consistence and feasibility of the content (internal
validity); capacity to corroborate the information through other sources.
Also we are interested in the following criteria regarding the meta-analysis:
3) Revise as many studies as possible; locate them through objective and arguable
searches; do not initially exclude studies because of their quality and differentiate
and classify each study according to the effect of its results.

The joint consideration of the previous criteria makes up a new approach that we
have called qualitative meta-analysis. Its finality, as any other meta-analysis, is: “…the
formulation of theories able to explain the phenomena observed in different
researches” (Bisquera 1989, p. 247-252); the difference in this case lies in the use of
criteria which are typical from an interpretative approach.

Consequently, we call didactical analysis of a specific mathematical topic to the
global methodological procedure that integrates and relates, by following a sequential
process and according to the criteria of the qualitative meta-analysis, information
related to the object of study coming from five basic areas: History and Epistemology,
Learning and cognition, Phenomenology, Teaching and curricular studies and
Sociocultural aspects. The sequenced process has the following stages:

First stage: Primary revision of the information in every area, following the process
in figure 3 and the following steps: a) analysis and classification according to the
European Research in Mathematics Education I.II: Group 7 254

established criteria; b) gathering the most important data; c) analysis of the
relationships among the data; synthesis and conclusions; d) conjectures and research
priorities in every area; e) assessment of each one of the area’s revision.

Second stage: Analysis of the relationships among central areas according to the
diagram in figure 2 and the following process: f) study of the relationships starting from
the information of sections c), d) and e) in each area; g) conclusions; h) conjectures and
priorities; aspects to research; i) general results and assessment.

Sequenced process on the information sources
for the Didactical Analysis
A phenomenon of interest

Search process
History and Epistemology of
Mathematics: Mathematics:
results and construction processes and meanings

sociocultural relationships
epistemological and through history
phenomenological historical analysis
analysis phenomenological analysis

New perspectives basic information
complementary Deeper comprehension
Transcription of basic information to an
environment under didactical purposes
Enculturation of information

Institutional sociocultural individual
relationships relationships relationships

programs teaching applications needs pupils teachers
textbooks classroom

Fig. 3: Integrating information in the first stage of the Didactical Analysis

Figures 2 and 3 describe the basic elements, their position in a sequenced process,
the main sources of information and the kind of analysis to be done. Throughout the
process data are compared and a global synthesis is carried out. As a consequence,
research priorities are established, theories and models are built and empirical and
European Research in Mathematics Education I.II: Group 7 255

experimental approaches are designed. The didactical analysis processes, analyses and
synthesises information coming from different fields linked to one another by its object
of study, giving a synthesis that let the detection of limitations of previous works and
properly organizing the future development of the research. The technique used bears
in mind the complexity of the field as well as the plurality of approaches and results we
can find in the scientific literature.

5. References
Bishop, A. J. (1992): International perspectives on research in Mathematics Education. In D.A.
Grouws (Ed.). Handbook of Research on Mathematics Teaching and Learning, 710-723. New
York: MacMillan Publishing Company.
Bisquerra, R. (1989): Métodos de investigación educativa. Madrid: CEAC.
Davis, P. J. & Hersh, R. (1988): Experiencia Matemática. Ministerio de Educación y Ciencia. Madrid:
Labor S. A.
Fischbein, E. (1990): Introduction. In P. Nesher, J. Kilpatrick, J. (Eds.), Mathematics and Cognition,
1-13. Cambridge: Cambridge University Press,.
Glass, G. V.; McGaw, B. & Smith, M. L. (1981): Meta-analysis in Social Research. Beverly Hills,
CA.: Sage.
González, J. L. (1995): El campo conceptual de los números naturales relativos. Colección
Monografías de Investigación. SPICUM Universidad de Málaga.
González, J. L. (1998): Números naturales relativos. Colección Mathema. Granada: Comares.
González, J. L. & Flores, P. & Pascual Bonís, J. R. (1994): Epistemología y Educación Matemática. In
L. Rico, J. Gutiérrez, (Eds.), Formación científico-didáctica del Profesor de Matemáticas de
Secundaria, 25-39. ICE Universidad de Granada.
Harel, G. & Dubinsky, E. (1992): The concept of function. Aspects of Epistemology and Pedagogy.
Washington, D. C.: MAA.
Kilpatrick, J. (1994): Historia de la investigación en Educación Matemática. In J. Kilpatrick & L. Rico
& M. Sierra (Eds.), Educación Matemática e Investigación, 15-96. Madrid: Síntesis.
Ogawa, R. T. & Malen, B. (1991): Towards Rigor in Reviews of Multivocal Literatures: Applying the
Exploratory Case Study Method. Review of Educational Research, 61(3), 265-286.
Ortiz, A. (1997): Razonamiento Inductivo Numérico. Tesis Doctoral inédita. Departamento de
Didáctica de la Matemática. Universidad de Granada.
European Research in Mathematics Education I.II: Group 7 256

Puig, L. (1997): Análisis Fenomenológico. In L. Rico et al. (Eds.). La Educación Matemática en la
Enseñanza Secundaria, 39-59. Barcelona: ICE Universitat de Barcelona.
Rico, L. (1997): Los organizadores en el currículo de Matemáticas. In L. Rico et al. (Eds.), La
Educación Matemática en la Enseñanza Secundaria, 39-59. Barcelona: ICE Universitat de
Romberg, T. (1992): Perspectives on Scholarship and Research Methods. In D.A. Grouws (Ed.),
Handbook of Research on Mathematics Teaching and Learning, 49-64. New York: MacMillan
Publishing Company.
Scriven, M. (1988): Philosophical inquiry methods in Education. In R.M. Jaeger (Ed.),
Complementary methods for Research in Education. Washington: AERA.
Tymoczko, T. (1986): New directions in the Philosophy of Mathematics. Boston: Birkhäuser.
Vergnaud, G. (1990): Epistemología y Psicología de la Educación Matemática. In P. Nesher &
J. Kilpatrick (Eds.), Mathematics and Cognition, 14-30. Cambridge: Cambridge University
European Research in Mathematics Education I.II: Group 7 257

Bruno D’Amore1, Hermann Maier2
Universita Degli Studi di Bologna, Piazza Di Porta S. Donato 5, 40127 Bologna, Italia
Universität Regensburg, Fakultät far Mathematik, Universitätsstraße, 93040 Regensburg, Deutschland

Abstract: More and more researchers in mathematics education recommend
mathematical writing by pupils as an important activity besides their verbal
contribution to classroom communication. A written text in which pupils express their
own mathematical ideas in their own language, i. e. by use of words and formulations
which are within their individual active language competence and performance, will be
called here, according to Selter (1994) a ‘textual eigenproduction’ (TEP). After a brief
characterisation of TEPS and discussing their didactical functions, we present the
design and the results of case studies with 16 teachers from Italy and Germany, on how
far they are prepared to and experienced in work with TEPS in their mathematics
Keywords: textual eigenproductions, analysis.

1. Pupils’ Textual Eigenproductions

What are textual eigenproductions (TEPs), of what kind can they be and what are their
possible didactical functions?

1.1 Characterisation of Textual Eigenproduction

In every day mathematics classroom pupils have rather much to write. However, the
main part of their writing is restricted to noting steps of the solution process and results
of arithmetical or algebraic tasks, like calculating the value of number terms,
transforming terms containing variables, and solving equations. These formalised
protocols follow rather exclusively fixed algorithmic procedures and standardised
patterns of symbolic representation; in so far we do not call them TEPs. But protocols of
European Research in Mathematics Education I.II: Group 7 258

problem solving procedures can easily be extended to TEPs if the pupils – using all
means of language actively available to them, also formulations of their every day
language – produce a detailed description of how they consider the problem content, of
what their aim(s), ways, and measures of solution are – there should also be given
reasons and justifications for the last ones – and of their results. The TEPs arising from
such activity can be called ‘commented problem solving protocols’ (see Powell &
Ramnauth 1992).

Of course, TEPs must not remain restricted to these one type. Other kinds of TEP are:
• Commented problem solving protocols (as described above);
• reports about mathematical investigations (aims, steps and measures, results);
• detailed descriptions and explanations of mathematical concepts or algorithms,
• texts initiated by a specific situation demanding to communicate mathematical
facts and relations in written form;
• texts defining mathematical concepts, formulating hypotheses, arguments or
proofs in relation to a mathematical theorem.

Writing in the meaning of TEP may happen from time to time in the mathematics
classroom; it also can become a regular pupils’ activity. Waywood (1992), e. g., reports
about journal writing, Gallin & Ruf (1993) call the texts their pupils produce regularly
‘journey diaries’; the writing deal with ‘core (mathematical) ideas’ on which they are
expected to reflect, and which they are wanted to invent.

1.2 Didactical Function of Textual Eigenproduction

There are some reasons why TEPs should be introduced into pupils’ mathematical
classroom work:
• TEP stimulates the individual pupil to analyse and to reflect on mathematical
concepts, relations operations and procedures, investigations and problem
solving processes he/she is dealing with. Thus, he/she can arrive at more
consciousness, and a deeper mathematical understanding of them.
• TEP is able to improve pupils competence and performance in technical
language, since it leaves them time for a careful and reflected selection of
language means, and thus encourages them to make an active use of technical
terms and symbols (see Maier 1989a, 1989b, 1993 and Maier/ Schweiger 1998).
European Research in Mathematics Education I.II: Group 7 259

• TEP gives the individual pupil a chance to get control about his/her understanding
of mathematical issues by means of a reasonable and reflected feed back from the
teacher’s and other pupils’ side.
• TEP enables the teacher to assess previously and actually constructed knowledge
and understanding of mathematical ideas in a more detailed and a deeper way
than it would be possible on the base of common written tests, normally carried
out in the manner of non-commented problem solving protocols.

The use of TEP in the classroom and moreover the positive effect on the learning
process and its evaluation mentioned above are certainly not a matter of course. There
are many preconditions for success; we restrict ourselves here to mention two which
appear us of high importance:
• If pupils are wanted to produce texts which can give deep insight into their ways
of mathematical acting, thinking and understanding it has to be made sure that
they address their TEPs to someone who needs full information about the matter
written about. Usually they tend to imagine the teacher as the only addressee of
their writing, whom they insinuate already to know all they have to communicate
and just wants to examine their ability of coming up to his quite specific
expectations. Thus, they feel no need for a detailed and explicit description and
explanation. Motivations suitable to change the writers attitude to a role different
from the pupil’s – e. g., there may be used instruction like „Imagine you were a
father/mother, a teacher, ...“ (see D’Amore & Sandri 1996 and D’Amore &
Giovannoni 1997) –, to write (a letter) to a younger child or a classmate who
missed classrooms for reason of illness and should be informed about what has
been learnt in his absence, to write a diary, to design a poster for an exhibition, to
formulate a lexicon article, etc. Or the teacher can organise a particular
communicative situation in which, e. g. a pupil has to describe a geometrical
drawing in a manner that a classmate is able to reproduce the figure only on the
basis of this description. Sometimes it may be helpful to defamiliarise the
common problem solving situation by means of open or incomplete tasks (see
D’Amore & Sandri 1997, 1998).
• The teacher does not only need good ideas for an effective stimulation of the
pupils, but also for an adequate way of working with the produced texts. Above
all he has to be prepared and able to interpret and to analyse the TEPs carefully
European Research in Mathematics Education I.II: Group 7 260

and competently. It is, indeed, not easy to find out the mathematical concepts,
thoughts and ideas which are on the base of the particular text produced. It needs
not only high attention but also a lot of experience and possibly some training as

The main research interest of the project the design and results of which we are
reporting subsequently was on the second precondition. In how far do present teachers
make use of TEP and how experienced are they in their interpretation and analysis? Are
they convinced that TEP could be a help for their mathematics teaching, mainly for
evaluating individual pupils mathematical knowledge and achievement? In detail the
questions of research were as follow:
(1) Which instruments do teachers usually apply for assessing pupils in mathematics?
What is in the centre of their evaluation (knowledge or abilities), and in how far and
in what way is their assessment individually or collectively oriented?
(2) Do they know further instruments of assessment (they do not use themselves), and
what is their attitude towards them? Do they, in particular, at least know tep as a
possible instrument of assessment, and which is their attitude towards it?
(3) How do they interpret and analyse TEPs? Which is their particular point of view and
the effectiveness and competence of their interpretation?
(4) Is a first experience with the analysis of TEPs able to modify the teachers’ attitude
towards them as an instrument of assessment (and in what way)?

2. Teachers Interpretation of Textual Eigenproductions

In this paragraph we describe the design of our investigation before we are going to
report about results. The investigation is methodologically designed according to the
rules of an interpretative research paradigm and qualitative research methods (see Beck
& Maier 1993, 1996, Glaser & Strauss 1967).

2.1 Design of the Research

We carried out a sequence of three interviews with each of 16 mathematics teachers
from Germany and Italy.
European Research in Mathematics Education I.II: Group 7 261

In the first interview the teachers were requested to report about what is in the centre
of their attention when they assess their pupils’ mathematical competence in the
classroom, and which are the means they usually apply. In addition they were wanted to
tell about their knowledge of assessment procedures not applied by themselves, and
about eventual reasons for not making use of them. In case TEP was not mentioned until
then, the teachers were asked if they know it, which were their possible experiences
with it, resp. what do they imagine under this label and did they ever consider its
introduction into their mathematics classroom. The interviewing person tried to put all
these question in an open manner, to react in a flexible way to the teachers’ utterances
and to put more exact questions in case of unclear or insufficient replies.

Afterwards every teacher was presented TEPs which have been produced by (Italian)
pupils unknown to him/her, as an answer to e. g. the following instruction (about the
origin of these texts, see D’Amore & Sandri 1997, 1998): Imagine you were a father (a
mother) .... Your young child, 7 years old, learnt from somewhere that every triangle
has three heights and asks you: „Dad (Mum) what does that mean ?“ Nothing is more
intriguing than leaving young children’s questions unanswered; therefore, you decide
to give the following reply: …

The teacher was given texts produced by the pupils e.°g. the one written by Simona
(Scuola Media 2; 11 to 12 years of age) saying: „My son, yet you don’t know geometry,
but I will explain you what height means. Like you I and papa have a height which is
measured from the head to the feet; the triangles have one as well, but their height is
measured from the vertex, which is a little point, down to the base, which is like our feet.
Since they have got three little points (vertices), they have three heights because they
have three pairs of our feet. And since we have only one head and one single pair of feet
we have only one height.“

Having received these TEPs the teacher was requested to read and analyse them
carefully until next day, in order to find out as much as possible about the writers’
mathematical knowledge and competence in reference to the respective topic.

In the second interview every teacher gave his interpretation of the TEPs he/she had
been given the day before. He/she was expected to make comments on how far the
pupils did understand the task and how he/she estimates their answer, to describe
European Research in Mathematics Education I.II: Group 7 262

eventual error strategies or wrong thinking on the pupils side, and finally to give a
judgement about the pupil’s ability. Afterwards he was wanted to consent to the plan of
letting his/her own pupils write texts initiated by the same instructions the Italian pupils
had been administered.

Some of his/her pupils teps, carefully selected mainly after the criterion of diversity,
were handed in to the teacher in a typed form and he was asked to be prepared for
analysing them like the Italian texts a third interview.

In the third interview the teacher started to talk about his/her general impression on
the pupils teps, before he/she entered into a detailed interpretation of one text after the
other. At first he/she could not be certain about the respective author and, thus, was able
to analyse it without making use of previous judgement on the pupils mathematical
knowledge and ability. Afterwards, he/she tried to guess the respective writer’s name
and was then informed about it. Knowing the name, he/she could make additional
comments, mainly on the question of if he thinks to have learnt something about this
individual pupil he/she did not know before, and what.

Finally the third interview turned back to a general discussion about tep. In case the
project gave the teacher the first opportunity of experiencing it, the interviewer wanted
to know what he/she thinks about it as a means for getting information about individual
pupils mathematical competence and ability. In case he/she already knew tep before,
the questions were directed towards possible change in his/her attitude towards it. In
any case the interviewer wanted to know if the teacher intends to make use of tep in his
future mathematics education.

All three interviews with every teacher were audio-taped and transcribed for further
analysis. All together, they formed the data base for one case study on the research
questions mentioned above. The investigation started with two cases, which were
immediately analysed, in order to formulate, by way of comparison hypotheses and
theory elements. Afterwards the next two cases were selected and analysed for means
of further comparison and further development of the theory (theoretical sampling in
the meaning of Glaser & Strauss 1967). In that way, so far 13 cases were studied with
teachers as follow:
• 8 German secondary school teachers, 2 of grade 6 (pupils of age: 12), 1 of grade 7
European Research in Mathematics Education I.II: Group 7 263

(pupils of age: 13), 2 of grade 8 (pupils of age: 14) and 3 grade 9 (pupils of age:
• 8 Italian teachers, 1 from „biennio superiore“ (pupils of age 14 to 16), 5 from
„scuola media“ (pupils of age 11 to 14), and 2 from „scuola elementare“ (pupils
of age 6 to 11)

2.2 Some Results

At first we report about the teachers’ description of the criteria and the means of their
procedures of assessment in the first interview, and how they interpreted Simona’s TEP
in the second interview. Then we report about the teachers interpretation of their own
pupils’ TEPs in the third interview and their concluding comments on tep as an
instrument of assessment in their classroom.

Criteria and means of assessment
Asked about the means they use and the criteria they are oriented at when they are going
to assess their pupils’ individual achievement in mathematics, many teachers
emphasised to apply several methods, but most of them – Italian as well as German
teachers – referred to the same, namely:
• Direct questioning or verbal examination in the classroom;
• small written tests (referring to a small topic area), in most cases without
announcement in advance;
• bigger written tests (referring to a larger topic area) of a lessons time at least, in
most cases announced in advance

Typical statements are as follow:
• Italian teacher D: „No, there is not a single method I use. We can say that I often
administer verbal or written examinations, without announcement as well. From
that I have always many information.“
• Italian teacher A: „I use the classical methods: verbal and written examinations,
small tests.“ – „From time to time I also administer written tests without
announcement. Nevertheless, most times the pupils know, they expect, they know
• Italian teacher B: „No, ... what is the matter? I ask the pupils many questions
European Research in Mathematics Education I.II: Group 7 264

directly in the classroom. They know that and answer. I like talking to them. This
is a nice report given in trust. In case they did not understand they themselves ask
me. And then, every four weeks I administer a written examination and also small

These common means of assessment can be characterised as formal. In most cases
the pupils have to answer to direct questions or to solve standard problems. Thus their
answers resp. their partial or final results can be judged as right or wrong, according to
norms fixed by the teacher. Thus, the outcome of the assessment can be used for a
pretended possible ordering of all pupils of the class in behalf of their individual level of
achievement. The order can even be sectioned into a few steps, represented by marks of
in between 1 and 5 or 6. It is, briefly said, a normative evaluation of pupils achievement,
the main criterion of which seems to be a correct and quick reproduction of memorised
definitions and algorithms.

But there are a few exceptions. The German teacher K. points out as his main
instrument of assessment the observation of individual pupils, which seem him able to
„make weaknesses evident quickly, and can be released individually“. However his
observations are restricted to verbal contributions and to results of individual, partner or
group work in the classroom.

The German teacher S. also directs his attention to observation in classroom. He
organises individual work, in which the pupils work independently and he „can walk
through the rows and see, what kind of difficulties some pupils have, and help to
overcome them“. In contrary to K. he also demands individual pupils to work at the
blackboard and organises classroom periods in which the pupils have to work

The German teacher We sets up, among other means of assessments, brain-storming
at the beginning of every classroom. There, the pupils tell all they know about a given
topic. That way the teacher gets at least a kind of collective control on their competence.
His intention is „to get a holistic impression of pupils’ knowledge about facts and
formulas and about their ability to apply them in adequate situation as well“.

The German teacher K applies a really extraordinary mean of assessment: „I
European Research in Mathematics Education I.II: Group 7 265

provoke the pupils by very difficult or nonsense problems, which they try to solve in
groups. The group is intended to succeed in finding a meaningful solution, which is
realised as soon as all group members one after the other, also the slow learners among
them, is able to explain the solution to another pupil, who was not able to find a

The question if the teachers know means of assessment which, for some reasons,
they do not use themselves brought little concrete answers. Some teachers simply
confess not to know further ones. Others restricted themselves to say there are several
ones, but they all are not applicable. Nobody mentioned or even described such means.
Typical is the utterance of Italian teacher B: „Of course I know more than one, but I am
not sure, what you want?“ Only one teacher mentioned TEP spontaneously in this part
of the interview, this is the German teacher S.

The German teacher S calls tep a „mathematical composition“. There are some
reasons why he does not apply it: „Mainly because of many pupils from foreign
countries with heavy language difficulties I can not expect satisfying results. In
addition this kind of assessment demands an enormous expense of time, not only in the
classroom but also for revision.“ Nevertheless, he thinks its use meaningful in a class of
sufficient language skills.

Directly addressed with the question if he knows TEP, the German teacher We said
he had some experience with TEP and thinks it to be „an adequate mean of assessment“,
but for himself he restricts it to the topic of equations. In this case the use of TEP seems
him „particularly adequate, since due to the production of text the pupil has to build up
the logic of the task and he/she needs a fundamental understanding“. However, he
wants to distinguish between assessing and evaluating (marking) pupils achievement,
since he never would tep use for the latter one.

Similar utterances made the German teacher Gr who teaches at the same school as
teacher We. He also let pupils write teps which deal with solving equations, and he
admits that, so far, he „never let the pupils write on geometrical questions“, and that he
has „difficulties to imagine how that could look like“.

The German teacher Wö claims to know TEP and says „that he could well imagine to
European Research in Mathematics Education I.II: Group 7 266

use it as a means of assessment.“ But in case of his present class he „would have little
hope to come to a reasonable result“. The other German and all Italian teachers
confessed not to know TEP.

Teachers interpretation of Simona’s TEP
By two German teachers we received quite interesting analyses of Simona’s TEP.
Therefore, they shall be quoted in full extension:
• Teacher Gr: „I think up to this point it is not bad: ‘... since the height is measured
from the vertex and since the triangle has three vertices there is one height
possible from every vertex to the baseline’ up to the statements about the three
pairs of feet. From that point on I can no longer follow her.... well, she should
have said that the height is perpendicular to the baseline or something like that,
but I think this could be clarified in a talk. They do not just think of everything,
while having to formulate. However I think it could be said that in fact she did
understand a bit.... I also thought about how I myself had answered the question,
but even I would not be able to explain it to a seven year old child in a manner
that it would enable it to understand. For certain I could explain it
mathematically but in a other way probably with help of a plummet or so. It
would have to be explained rather manually, otherwise a seven year old child
will not understand. So to say there should be presented a real triangle and made
fall down a plummet, turn the triangle and make fall the plummet again and so on
etc. That way it can be seen that it always goes down vertically. But certainly
nothing can be achieved with words like ‘perpendicular’. I can tell you quite
seriously that I myself had some difficulty with this task. Therefore, I think that
probably many pupils go back to what they learnt in the classroom and less enter
the question how it could be explained to a child.“
• Teacher S: „Yes, I have to say, not bad. If the child really is of age seven the
answer fits to a certain extent. In parts the answer is not bad; I would only say,
that the father or the mother could provide an illustration for the child. And
about the height: is it allowed to say we have one height and the triangle has
three? I would see that rather a bit different. I would not make such strong link to
the height of man. Okay, it can be said, there is some relation, but the link of the
one height of man and the three heights of the triangle is not so good. But in
principal, for a seven year old child such an explanation, why not. It should be
European Research in Mathematics Education I.II: Group 7 267

said in addition how this height is to be located, that it is the shortest tie from
exactly this point to the opposite line. This should be included.“

These analyses, given by the two teachers who already had some experience with TEP,
are untypical for at least the following reasons:
• Going into the details of Simona’s text, they try to give a complete description of
her concept and ideas about height indicated in the text, and they compare them
with their own – as the firs text shows, possibly questionable – mathematical
knowledge about this issue.
• As far as they evaluate the text they do it in a differentiating way.

This is different in the case of the following statements:
Italian teacher B: „Oh yes I have been amused. This answer hits the topic well, an
answer of high level, since she is able to explain this thing to child of 7 years, that
means that in her mind she has perfectly understood what the heights of a triangle are...
You have also to see how she tries her best to say it in a good manner.

German teacher We.: „I think the pupil did fundamentally understand the problem.
She explains this in a nicely visual way and in a manner that someone without previous
knowledge could get a good imagination. I would be happy about this answer.“

German teacher K.: „I find this Simona excellent. The answer has to be read
carefully: ‘A triangle has three tops.’ And in a pupils mind normally the triangle is
normally situated like that: ‘The point on top the feet at the bottom.’ She simply turns
the triangle around. A triangle has three heads, compared with man. In case this is
really a child, it must be of really high intelligence. In relation to a seven year old child
the question is actually completely answered. Hits the point without blab. Very visual. I
am quite exited.“

Teachers’ interpretations of their own pupils teps
The interpretations typically given by nearly all teachers in reference to Simona’s TEP
and to the TEPs of their own children – in case of Italian and German teachers alike –
hardly differed from each other.

They shall be characterised at hand of five TEP examples as follow:
European Research in Mathematics Education I.II: Group 7 268


„The triangle has only one height and a
b h a baseline by means of which area and
circumference can be calculated.“

(II) C

“A triangle has always one height and this is
always from the baseline to the point C. The
height always starts perpendicularly from the
baseline to the point C.”

A g B
(III) (green point)
„Here I’ve drawn a triangle and when I now
draw a line from the green corner downwards
to the straight line this is called height. When I
now draw a line from the orange corner
upwards and from the green corner draw a line
to the side the line upwards is also called
height. When I do the same from the pink corner
and from the green corner as well, the line
upwards is again called height.“
(pink point) (orange point)

(IV) „Triangle is a plain figure, this triangle has one height, and since the triangle can be turned
around it consequently has three heights.“

(V) C

„The first height goes through the vertex C. It
stands straight on the opposite segment. The
second height goes through the vertex B. It
stands straight on the opposite segment. The
third height goes through the vertex A. It stands
straight on the opposite segment.“

European Research in Mathematics Education I.II: Group 7 269

a) Evaluating, not descriptive interpretation: Above all the teachers’ interpretations
were strongly characterised by evaluating more than describing utterances. The
teachers seemed not so much interested in what the individual pupil really thought
about the topic in question, i. e. in the mathematical concepts and mathematical
knowledge indicated by the text. They rather draw direct conclusions on the pupils
general level of achievement. Some German teachers even felt able to fix this level by
attributing exact marks to the text author. For example TEP (I) was commented by the
author’s teacher as follows: „Perhaps I would say this is a pupil from the middle field.
That about baseline and the height for the calculating the area is correct, but it is not
asked for the circumference. Here he has mixed up something.“ After being told the
pupils name: „Yes, he really tends more to the better ones; so to say he tends to two.“

TEP (IV) was very briefly commented by the teacher, saying: „Since one is able to
turn the triangle around. Well, I think, probably, he may have thought of the right thing,
any way, because the turning seems to be a kind of correct approach.“

What the teacher not realised was that pupil who produced TEP (I) evidently regards
lines in the triangle not as autonomous (geometrical) objects. The words „baseline“ and
„height“ represent in his mind magnitudes which are needed to calculate the triangle
(area and circumference). That means that he interprets them functionally under an
elaborative aspect. The triangle „has“ the baseline and the height so to say exactly for
that reason. Thus, it has to be seen as consequent not to make further differences
between both. In calculation both have the same function; there is no need for
differentiation. That goes so far that the pupil evidently uses the labels „baseline“ and
„height“ synonymously, regards the baseline as the second height and talks about a
„third height“ although so far he had only mentioned one. The instruction evoked in the
pupils mind evidently rather an algebraic than a geometrical imagination. Even the
word „triangle“ is used by him less as a name of a geometrical object than as a stimulus
for selecting a certain algebraic procedure.

This is quiet different in case of text (IV) in an expressively conceptual approach its
author calls the triangle a „plane figure“; i. e. she attributes it conceptually to the
category of plane geometrical figures (polygons). The pupil also shows a quite clear
intermodal conceptual imagination about height; in her opinion this is a line which only
exists in a triangle one side of which is horizontal, i.e. parallel to the top and bottom
European Research in Mathematics Education I.II: Group 7 270

margin of the sheet. This is since the height has to be vertical, i. e. parallel to the right
and left margins of the sheet. In spite of this quiet narrow concept of height the text
author succeeds to give a reason for the triangle having three heights. She imagines to
turn the triangle in the plane so that one after another each of the three sides once
becomes horizontal. That way it has a height for each of these sides, what means in total
three heights. (It seems interesting that in this case a line of a triangle can loose its
property to be a height when its position changes. On the other hand the line gets this
property back at any time when the particular position is restored. This is sufficient to
make the existence of three heights argumentatively certain.

Possibly the teachers followed their habit to use written products from the pupils’
side for assessing their mathematical achievement according to class internal norms.
Thereby they did not so much follow what is called a central tendency. Rather they
tended to evaluate the text (in reality the pupils) in a strongly dichotomic way. They
appeared to them as either totally good or absolutely inadequate or bad.

The criteria on which the teachers based their valuation were mainly about
• how much a pupil came up to the norms of a formal mathematical language (in
many cases, correct mathematical statements given in the pupils own language,
were not accepted),
• how near the pupils utterances were to the own teachers content and way of
presentation in the classroom; they were seen more positively in case they
showed similarity with contents presented or the language used by the teacher,
and original ideas from the pupil side were, at the most, classified as
„interesting“, but not highly appreciated.)

For the criterion mention last the interpretation of TEP (II) by the teacher gives an
interesting example: „This answer possibly has already something to do with the
Pythagorean Theorem. Since we already have started with it, and, therefore, it may be
that this plays a role in the answer. With the perpendicular line where we then
sometimes have calculated the baseline. This is possible since we particularly went
down starting from the vertex C. I would rather attribute this to a good pupil.“ And
confidently when told the pupil’s name: „Yes, Stefan has a two.“
European Research in Mathematics Education I.II: Group 7 271

b) Unequivocal, not open interpretation: As a surprise, evidently the texts appeared
unequivocal to nearly all of the teachers. They felt able and showed no doubt to decide
exactly about the authors’ thinking and even about their particular level of
achievement. None of them generated more than one particular interpretation about
the author’s mathematical concepts and ideas. Example:

In text (III) the height which goes from the „green“ point to the „straight line“
(means the horizontal drawn baseline) is relatively clearly to be identified. But how
shell we interpret the following sentence „When I now draw a line from the orange
corner upwards, and from the green corner draw a line to the side the line upwards is
also called height.“ There are at least two interpretations which seem of the same
probability which can be made clear by the following drawings.

“line to the side”
The pupil imagines a height line
“line upwards”
outside the triangle, perpendicular to “green corner”
an elongation of the opposite side.
“orange corner”

The pupil has a height line in this “line to the side”
mind which is parallel to the right “green corner”
and left margin of the sheet and runs “line upwards”
towards a parallel to the (horizontal) height
baseline through the „green“ edge.
“orange corner”

The teacher, however gave a very simply analysis like that: „From the formulation it
could be Johann B. he is between two and three and tends to make careless errors. He
does formulate it completely, I am rather pleased with it, since I myself work much with
coloured chalks at the blackboard. The drawings even become more understandable
you can better realise the relations. Really, I like that rather well.“

c) Global, not detailed interpretation: It could be observed that the teachers did not
analyse the texts word by word, sentence by sentence in a detailed manner. Thus, they
were neither able to differentiate their valuation and to relate it to the mathematical
European Research in Mathematics Education I.II: Group 7 272

content, nor to describe the pupils mental processes resulting in their text in detail and
with sufficient exactness. In most cases the teachers analyses can be regarded as
superficial and incomplete. TEP (V), for example, got this interpretation: „I think that
pupil understood the task with the heights. He fixes the height exactly through vertex C
and through the opposite side; the other two as well. Certainly he does not say that they
have to be perpendicular to the baseline. But the answer is not that bad.“

In this text, indeed the teacher picked up a lot. But he could have mentioned that the
pupil does not say „through the opposite side“ but describes that each height „stands
straight on the opposite segment“ (German: „steht gerade auf der gegenüberliegenden
Strecke“). This can be seen as an exact equivalent from every day language for the
mathematical expression „is perpendicular to the opposite side“. In addition the text
author talks about „the first“, „the second“ and the „third height“. This is his way of an
indirect proof of the statement: „The triangle has three heights“.

In text (I) it should have been taken into account that the text describes „triangle“
and „height“, talks about „baseline“, „area“ and „circumference“ and that it also
mentions a „third height“, although until then it had only be talked about a single one.
In addition the author points out that he does not know anything about this third height.
Also TEP (III) speaks in the I-form. In contrary to (I) it relates excessively to a added
drawing in which the edges of the triangle are marked in colours. The pupil talks about
actions of drawing, where certain directions are attributed to lines: „downwards“,
„upwards“, „to the side“, „towards the strait line“. And it forms if-then-sentences: „If I
draw..., this upwards running line is called height.“

Teachers’ final comments on TEP
Many teachers were surprised when they compared their evaluation of the TEPs with
their previous meaning about the author. Many found excuses for „good pupils“ having
written „bad“ texts and „bad pupils“ having written „good“ texts and did not get the
impression that TEP could help them to learn more about their pupils mathematical
competence and thinking. Few, however, ended up with exactly that conviction and
promised to use the instrument regularly in the future.

If the didactic function of teps, described under 1.2, ought to become reality, it seems
European Research in Mathematics Education I.II: Group 7 273

absolutely necessary that the teacher makes use of them in an adequate manner. This
means that he provides regular training for the pupils in text writing, and that he is
prepared and able to interpret and analyse the texts in a descriptive rather than
evaluating way, in detail and completely and not in quite a general and selective
manner, conscious that every text is open for different interpretation. Evidently this
ability is not a matter of course; it has to be learned and has to be trained.

In addition the teacher has to be convinced that his teaching and his organisation of
learning processes can have, and normally has different effects on the individual pupils
side. These differences are not only quantitative ones, i. e. different pupils pick up more
or less from the teachers offers, but are in the same way of deeply qualitative kind. This
means that the pupils construct qualitatively different concepts, knowledge and
mathematical thinking. The teacher should be eager to learn about this differences and
to draw consequences from it for the planning of his classrooms.

Making pupils write TEPs and appreciating this activity as a stimulus for learning
and a means of assessment has to be based on a particular philosophy of mathematics
and mathematics learning. It demands a particular opinion about what it means for the
pupils to do mathematics.

3. References
Beck, Ch. & Maier, H. (1993): Das Interview in der mathematikdidaktischen Forschung. Journal far
Mathematikdidaktik 13 (2), 147-179.
Beck, Ch. & Maier, H. (1996): Interpretation of text as a methodological paradigm for empirical
research in mathematics education. A. Gagatsis & L. Rogers (Eds.): Didactics and history of
mathematics. Thessaloniki.
Beck, Ch. & Maier, H. (1994): Zu Methoden der Textinterpretation in der empirischen
mathematikdidaktischen Forschung. In H. Maier, J. Voigt, Verstehen und Verständigung im
Mathematikunterricht – Arbeiten zur interpretativen Unterrichtsforschung, 43–76. Köln:
D’Amore, B. & Sandri, P. (1996): „Fa finta di essere...“. Indagine sull’uso della lingua commune in
contesto matematico nella scuola media. L’ensegnamento della matematica e delle scienze
integrate 19A (3), 223–246.
D’Amore, B. & Giovannoni, L (1997): Coinvolgere gli allievi nella costruzione del sapere
European Research in Mathematics Education I.II: Group 7 274

matematico. Un’ esperienza didattica nella scuola media. La Matematica e la sua didattica (4),
D’Amore, B. & Sandri, P. (1998): Risposte degli allievi a problemi di tipo scolastico standard con un
dato mancante. La Matematica e la sua didattica (1), 4-18.
Davidson, D. M. & Pearce, D. L. (1983): Using Writing Activities to Reinforce Mathematics
Instruction. Arithmetic Teacher 35 (8), 42–45.
Gallin, P. & Ruf, U. (1993): Sprache und Mathematik in der Schule. Ein Bericht aus der Praxis.
Journal far Didaktik der Mathematik 14 (1), 3-33.
Glaser, B.G. & Strauss, A.L. (1967): The discovery of Grounded Theory. Strategies for Qualitative
Research. New York: Aldine.
Kasper, H. & Lipowsky, F. (1997): Das Lerntagebuch als schülerbezogene Evaluationsform in einem
aktiv-entdeckenden Grundschulunterricht – Beispiel aus einem Geometrie-Projekt. In
J. Schönbeck (Ed.), Facetten der Mathematikdidaktik, 83-103. Weinheim: Dt. Studienverlag.
Maier, H. (1989): Problems of Language and Communication. E. Pehkonen (Ed.), Geometry –
Geometrieunterricht, 23-36. Helsinki: University.
Maier, H. (1989): “Conflit entre langage mathématique et langue quotidienne pour les élèves”.
Cahiers de didactique des mathématiques (3), 86-118.
Maier, H. (1993): “Demande che si evolvono” durante le lezioni di matematica". la matematica e la
sua didattica (2), 175–191.
Maier, H & Schweiger, F. (1998): Mathematik und Sprache. Zum Verstehen und Verwenden von
Fachsprache im Mathematikunterricht. (in print).
Miller, L.D. (1992): Teacher benefits from using impromptu writing prompts in algebra classes.
Journal for Research in Mathematics Education 23 (4), 329-340.
Phillips, E. & Crespo, S. (1996): Developing Written Communication in Mathematics Through Math
Penpal Letters. For the Learning of Mathematics 16 (1) 15-22.
Powell, A.B. & Ramnauth, M. (1992): Beyond Questions and Answers: Prompting Reflections and
Deepening Understandings of Mathematics using Multiple-Entry Logs. For the Learning of
Mathematics 12 (2), 12–18.
Selter, Ch. (1994): Eigenproduktionen im Arithmetikunterricht der Primarstufe. Wiesbaden: Dt.
Waywood, A. (1992): Journal Writing and Learning Mathematics. For the Learning of Mathematics
12 (2), 34–43.
European Research in Mathematics Education I.II: Group 7 275

André Rouchier
CREHSTO -Universite d’Orléans, IUFM Orléans-Tours, 72 Rue du Fbg de Bourgogne, 45044 Orléans, France

Abstract: During the past thirty years the theory of didactical situations developed by
Guy Brousseau and other people was at the centre of the identifications of the main
characteristics of didactical systems as far as they are set up to teach mathematics.
Following the great trends of the theory in its historical and notional development,
from the first modelisation of the mathematization process to the identification of the
crucial role of institutionalization, we focus our attention on the more recent questions
calling for evolutions as new fields for research: regulation of didactical systems,
organization of sets of problems for teaching, relationship between teaching devices
and concepts of the theory in various contextual environments such as teachers
training and the use of so called new technologies of information.
Keywords: didactical situations, institutionalization, regulation.

1. Introduction

The relations between teaching and learning are at the centre of concerns for all of those
involved in the field of mathematical education, as much through the everyday action in
the class as for an action within less immediate range which leads on to the construction
of a curriculum proposal.

Il is also a central matter for research, which could plan as a reasonable purpose to
provide one or several theories giving an account of associated phenomenons. From the
famous Socratic dialogue of Menon to the most recent models as the Bereiter one
(1985), we all know many ways to give an account of the fundamental paradox of
teaching-learning. We are going to develop briefly here one of the most powerful
theories for the study of this matter in the mathematical field, the theory of didactical
situations. It was built on the identification of the founding role of the didactical
European Research in Mathematics Education I.II: Group 7 276

intentionality and on its realization through the a-didactical/didactical dialectic. We
will bring to the fore the central role of institutionalization in this process and we will
try to identify how the theory of didactical situations allows to undertake the study of
matters the teaching of mathematics comes up again.

2. The Classical Components of the Theory of Didactical

At present we know two theories which get organized on the identification of facts and
phenomenons of a didactical sort. These two theories, the one of didactical situations
and the anthropological theory, are structured on the following fundamental principles:
the mode of social realization of the didactical intentionality towards a learning of
knowledge, social practices or techniques considered as stakes, are inseparable from
the epistemological and functional features of these knowledge.

Studying the different forms these modes of realization take in the context of a
personal plan or in the context of a social one - which is the case of mathematics -,
studying the restraints they have to deal with, the variables structuring them and
determining the space for the didactic action as well as the role, places and
responsibilities of people acting, teachers and pupils, represents the central object of
research in didactics.

The production and the construction of mathematical knowledge, as well as their
transmission, follow equivalent structuration roles. The mathematical knowledge, in
the broadest sense of the term, cannot do without didactics, as much in its development
as in the forms through which its transmission and its teaching are achieved. Therefore
we can consider that there is an autonomy and a specificity of the didactics, especially
concerning mathematics. The psychodynamic process from Dienes, proposed in the
beginning of the sixties as a pattern for the construction of mathematical knowledge
showed a certain amount of restrictions which did not allow him to suitably convey
characteristics of the didactical interactions. It’s greatly to Guy Brousseau’s credit that
he proposed another modelling allowing to overcome the difficulties of the former
pattern. The basic idea was this : a mathematical expression process consists of several
European Research in Mathematics Education I.II: Group 7 277

steps and is based on specific games for which objects are brought into play, these
objects being necessary for the realization of this process and being objects of
knowledge : game of interaction with an environment with regards to an action, game
of interaction with an environment especially structured for the communication with
regards to formulation, game of interaction with an opponent (which may include the
adjustment of decision rules) to sort out statements from their trueness character and to
determine those having a trueness character from those definitely not having it. The
study of basic situations corresponding to the dimensions of action, formulation and
validation produced significant work and allowed in certain domains, wether to let
some questions come into the didactic field whereas they did not seem to be relevant
through look of a suited problematic, or to renew the approach of certain problems.
Let’s quote particularly and respectively the research work of J. Perez (1983) about the
creation of a code at kindergarten, the work of N. Balacheff (1988) about proof and
demonstration, and the numerous situations G. Brousseau and his students built and
proposed in the context of the COREM. Hence the theory of situations formed itself
into a theory of learning derived notably from the cognitive psychology, the one of
learning by adaptation. In this respect, it’s important to notice that the transfer made by
the theory of situations has got as a result (as much as an object) to move the place of
questioning from learning itself to the conditions of learning; Then it will be possible to
come up to the theorization of the teaching facts without the prerequisite of a theory of

Let’s resume these ideas to go further on. These are ideas structuring the
thematization peculiar to the theory of situations. Mathematics is the product of specific
games having dimensions of action, information and veracity. To learn mathematics,
one has to meet these games, hence to practice them. These basic games exist in the
familiar environment of any individual or young child. Some of them are timely
organized by adults, some are not. The later ones are called no-didactical situations, the
former ones are called didactical situations (or with didactical components).
Unfortunately, in most of the cases, what is at stake in these situations is not the
mathematical objects or knowledge. Only the elementary numerical knowledge as the
one of the chain of numbers are acquired in this way, especially in the family circle. The
acquisition of more sophisticated mathematical knowledge can only be very rarely
produced by circumstances when no didactic intentionality external to the individual
European Research in Mathematics Education I.II: Group 7 278

appear. In other respects, the situations are not all equivalent as regards to the way a
knowledge is brought into play, and they are even less equivalent when it is a matter of
taking this knowledge at stake. The assumption is made that some of these situations
have got a special generic value, these are fundamental situations. The determination of
fundamental situations of mathematical knowledge is a central issue for the research in
didactics. Two examples especially emblematic of situations having the required
characteristics are to be found in Brousseau (1997). The effective realization of these
situations, or of situations possessing rather good characteristics in a school context,
introduces the non didactic in the heart of the didactics or of a didactical project. In that
case, we can speak of an a-didactical situation. The a-didactical situations, as regards to
their conception, to the realization and to their leading (by the teacher), involve a high
level of requirement. Actually, within their principle itself comes the ability (their
ability) to allows the appearance, in a determined form, of a new knowledge, only by
the interplay of interactions led by the situation and without any direct didactical
pressure from the teacher.

3. Knowledge and Institutions, the Particular Case of

Mathematics would not be the product of a spontaneous generation. Actually, some of
their ways of existence depend on the social devices and on the sets of practices in
which they fit. There is a form of universality of the mathematical theories, and this up
to a certain point, as showed by the existence of the intuitionist program. However,
techniques themselves are not always universal. We actually know, and the
ethnomathematics work confirms it, that stable and relatively specific sets of
mathematical techniques exist, and spread in organizations and/or fields of practices for
which they fit needs. M. Douglas (1985) especially studied this phenomenon in “How
institutions think”. Then we will generically define as an institution any specific social
grouping, operating on the symbolic mode and for which a knowledge is at the principle
of its constitution. Hence institutions and knowledge (of the institution) co-define each
others. The mathematical objects, the mutual relations they keep up, their modes of
designation, the calculations associated to them, the problems and questions they allow
the study of, appear in numerous institutions, from home economics to astronomical
European Research in Mathematics Education I.II: Group 7 279

calculations, or the one of mathematical science itself, the one of professionals. Each of
these institutions has got specific needs, and must therefore learn the mathematics
useful to it, as well as for any people entering the institution, hence wishing to have,
with the other people from the institution, relations consisting in having common
mathematical practices, the ones of the institutions. From the point of view of the
transmission of mathematical knowledge, we will consider as phenomenons of the
same hand the ones concerning the individuals and the ones concerning the institutions.

The relations an individual, a group of individuals or an institution has with
mathematics, relations updated in practices where mathematics can be observed
(techniques, statements, forms of proofs) are not always explicitly pinpointed by their
actors as knowledge. In this way, one can calculate areas without being able to express
some properties of the notion of area (not having studied it or not having to study it).
One can use mathematical techniques enabling to solve division problems by
calculating a quotient and a remainder without to be in a position to identify these
techniques as being a matter for division. These techniques, even of they have been
learned, wether by mutual contact, wether in a non didactical or a didactical context,
appear as knowledge. They will be converted in learning by institutionalization,
becoming autonomous from the conditions of their emergence, de-contextualizad. “Les
instruments culturels de reconnaissance et d’organisation des connaissances sont des
savoirs, objets d’une activité spécifique des institutions, ou d’une activité d’institutions
spécifiques”, G. Brousseau (1995). The mathematicians’ community, by its permanent
work of reorganization of mathematics, and especially the most recent mathematics,
performs institutionalization.

A didactic theory of institutionalization has been studied by Rouchier (1991). A
general outline of institutionalization has been brought out by action (of
institutionalization) on an “acted” situation, in which knowledge mainly as knowledge
in action (situated knowledge) have been involved by the subjects. These results led to
the start up three series of work. The first ones are directed towards the explication and
the realization of this outline in specific contexts, relating to the objects and stakes of
knowledge connected to these objects. The research on the didactic memory of teachers
(J. Centeno (1995), the studies related to the structuration of the milieu (C. Margolinas)
can be linked with this first series.
European Research in Mathematics Education I.II: Group 7 280

The second ones are directed towards the study of the relations between the notion
of condensation as proposed by Arzarello, Bazzini and Chiappini (quoted by T. Assude
(1994)) and institutionalization work from T. Assude (1995), especially from the angle
of a-didactic theory about conceptualization.

The third ones are directed towards the modes of the actual realization of this
transformation of the “acted” situation and the ways students and teachers are involved
as “ cognitive” actors of the whole process. Among the various studies that were done
in the field, the work of G. Sensevy (1994) is one of the more spectacular. He put in
evidence, in the field of teaching-learning of fractions in the elementary school, the
crucial role of students’ writing about their own knowledge and the debates these
writing permit to set up. Theses debates are centred about crucial questions such as :
“What can be the signification of a fraction greater than one ?” coming from the
students and discussed by the whole class.

4. New Objects, New Problems

We are going to present now three series of questions, of problems being the concern of
the didactic sphere and calling for new developments of the theory of situations. As
questions, they are not rigorously truly new, but a matter of fact the actuality of some of
them is modified, and moreover we have theoretical and experimental means at our
disposal to suitably problematize them.

The first series of questions is associated to the stability of the didactical system. We
can actually express as a general principle, due as much to necessity (in the sense of
being necessary) as to the observation of its empirical nature, that any didactic system
cares about upholding its stability. We will not define here a notion of stability for the
didactic systems and we will accept to endow it with an equivalent meaning to “resist
the disturbances”. These disturbances are numerous, potentially as well as effectively,
and their characteristics depend on the system being observed. For instance, in a
classroom, some mistakes made by the pupils are going to request corrections only
appealing to a simple individual treatment for which the teacher intervenes with only
one pupil, and this will not be an important disruptive factor for the didactic system. On
the other hand, some other, heavier or more significant mistakes as regards to the past or
European Research in Mathematics Education I.II: Group 7 281

the present of teaching, need much more important interventions involving among
others, injunctions to re-learn a lesson, explanations, complementary exercises... In
other cases, the didactic system in its whole is reacting, integrating or rejecting a
curricular modification, which is a matter for a similar phenomenon. G. Brousseau
(1996) brought to the fore the fundamental structural role of regulation, in particular
about the understanding of the teachers’ role. It’s a fundamental characteristic of his
activity. G. Brousseau proposed an aid to study the regulation’s instruments and the
teacher’s strategies : the changing of contract. Hence we can consider a typology “a
priori” of different communication contracts whose stakes are knowledge, some of
them being far beyond the mere communication. The principle of classification of these
contracts consists in identifying the respective roles and responsibilities of the
transmitter (in general the teacher) and the receiver (in general the learner) concerning
the devolution of the exercise of an operating responsibility towards knowledge. The
dynamics of teaching-learning is managed by changes of contract occurring during the

The second series of questions is linked to the structuration of the didactic system, in
the sense of a dynamic sharing of sets of problems, activities, terms and symbolic
forms. This listing is not complete and the whole cannot strictly speaking stand for a
definition. To be operating from a didactic point of view, these sets of problems have
got internal and external properties, enabling them to ensure their function, which is to
allow teaching and learning. Following Y. Chevallard (1996) we can designate these
sets as organizations.

Some of these properties have an epistemological dominant characteristic. They
contribute to ensuring the appropriateness of knowledge and learnings, whose
examined set allows their construction, with elements of mathematics in their double
reference to the scholarly learnings and to the reference practices. In this scope, we can
feel interest for the various environments in which some questions, some problems and
therefore some objects, knowledge and techniques will be called for and contribute to
founding the meaning of the aimed knowledge. Hence for example the notion and the
problem of scales enable to train to work about the notion of ratio in numerical domains
where one element of the ratio is a large number.
European Research in Mathematics Education I.II: Group 7 282

Other properties have a didactic dominant characteristic. On this account, let’s
quote the rate of use of some techniques, the order of appearance of problems and
exercises, the representivness of some subsets of problems, the resumptions (of terms,
of techniques, of problems). This listing is obviously not exhaustive and numerous
studies are still necessary to ensure good identification of these functionally relevant
properties of these sets. We must underline that the proposal worded here is not made
without knowing the learnings many participants of the educative system might have
developed during their practices about these sets of exercises (among others, teachers
themselves and scholarbooks writers).

The third series of questions will call for the development of new and specific means
of study, as much on the theoretical level as on the empirical one, the one of
confrontation with contingency. It’s a matter of studying, from a double view point of
modelling and determining scopes and means for action, the connections between the
effective school plans (as defined by constraints of the system which are specific
neither to mathematics nor to strictly speaking didactics, but referring to general
determiners of school organization) and the didactic systems which can, and have to,
take place inside these plans. Three series of considerations bring these questions as a
current issue.

The first one is linked to the progresses brought by the results of research in the field
of mathematics teaching, and among others the one produced by the thematizations
having a didactical characteristics (theory of situations and anthropological approach).
Innocently speaking, the functioning of elements stemmed from research in a standard
school context need to have them undergo specific adaptations even if the can rely on
teachers’ know-how.

The second series of considerations is linked to the recent diversification of the
information media and the auxiliaries of the mathematical activity which are the
mathematical softwares (Derive, Maple, Cabri Geometre) and the pocket calculators.
They lead to evolutions in the domain of teaching and learning devices, but these
evolutions remain restrained by the putting into practice of the didactical intentionality,
founding and characterizing the teaching and learning systems at a scale in keeping
with the scopes relating to these systems. This didactical intentionality is the one which
European Research in Mathematics Education I.II: Group 7 283

gives structure to the didactic systems and determine their invariants and their

The third series of considerations is associated to the new requirements of teachers
training, being an initial training or an in-house training. All of us have got a practical
knowledge of the difficulties any participant (trainer or trained) operating inside our
training systems might encounter, as it is an important part of our professional activity.

Instead of keeping in sterile confrontation such as theory versus practice, still
cluttering too much our ways of thinking in this domain, it would be proper to substitute
ways to generate the action of teaching and its relation to the didactics, bearing a real
operatory value and being able to fit in with both the brief period of time of the initial
learning and the longer periods of time of the in-house training; It is a matter to the
study of which research can bring an amply significant contribution through adapted

5. Conclusion

In this paper we choose to took at things from a general point of view to sustain the
debate which should first find its place in a set of themes dealing with the relationship
between theoretical problematics and questions about teaching. From this view point, it
seems more fruitful to detail a few central questions so that the issue of their relevance
and their signification in the context of research in mathematic teaching could be
brought up for discussion. Some of these questions already are being studied, set from
the point of view provided by the theory of situations. Some others still don’t have
completely found this place allowing them to be studied from this viewpoint. The work
this paper calls for might produce significant evolutions about relevant objects and their
mutual relations. Whatever the case, this can only be positive as much for the questions
themselves as for the chosen theoretical problematics.
European Research in Mathematics Education I.II: Group 7 284

6. References
Assude T. (1994): Quelques principes de l‘écologie didactique des savoirs, condensation et densité de
formes de savoir, Séminaire DidaTech, Université J. Fourier, Grenoble.
Assude T. (1995): Condensation et institutionnalisation: poser le problèmes et questions ouvertes.
Séminaire DidaTech, Université J. Fourier, Grenoble.
Balacheff N. (1988): Une Étude des processus de preuve en mathématique au Collège, Thèse,
Université J. Fourier, Grenoble.
Bereiter C. (1985): Towards a solution of the learning paradox, Review of Educational Research, vol
55, no. 2, pp 201-226.
Brousseau G. (1997): Theory of didactical situations in mathematics, translated and edited by N.
Balacheff, M. Cooper, R. Sutherland, V. Warfield. Kluwer: Dordrecht.
Brousseau G. (1996): L’enseignant dans la théorie des situations didactiques, Actes de la VIIIième
école d’été de Didactique des mathématiques, R. Noirfalise, M.-J. Perrin-Glorian (Eds.), IREM
de Clermont-Ferrand.
Centeno J. (1995): La mémoire didactique de l’enseignant, Thèse posthume inachevée, texte établi
par C. Margolinas, LADIST, Université de Bordeaux I.
Chevallard Y. (1996): La fonction professorale, esquisse d’un modèle didactique. Actes de la
VIIIième école d’été de Didactique des mathématiques, R. Noirfalise, M.-J. Perrin-Glorian
(Eds.), IREM de Clermont-Ferrand.
Comiti C., Grenier D., Margolinas C. (1995): Niveaux de connaissances eu jeu lors d’interactions en
situations de classe et modélisation de phénomènes didactiques. In G. Arsac G, J. Gréa, D.
Grenier, A. Tiberghien (Eds.), Différents types de savoirs et leur articulation. La Pensée
Sauvage, Grenoble.
Douglas M. (1986): How institutions think. Syracuse University Press
Margolinas C. (1995): La structuration du milieu et ses apports dans l’analyse ˆ posteriori des
situations. In C. Margolinas (Ed.), Les débats de Didactique des Mathématiques. La Pensée
Sauvage, Grenoble.
Perez J. (1983): Construction d’un code de désignation â l’école maternelle, IREM, Université de
Bordeaux I
Rouchier A. (1991): Étude de la conceptualisation dans le système didactique en mathématiques et
informatique élémentaire: proportionnalité, structures itérativo-récursives, institutionnali-
sation, Thèse, Université d’Orléans.
Sensevy G. (1994): Institutions didactiques, régulation, autonomie, Une Étude des fractions au cours
moyen, Thèse de l’université de Provence, Aix-Marseille.
European Research in Mathematics Education I.II: Group 7 285

Nada Stehlíková
Department of Mathematics and Didactics of Mathematics, Faculty of Education, Charles University, M.D.Rettigove 4,
116 39 Prague 1, Czech Republic

Abstract: The article focuses on the methodology of research within the Prague
Seminars of Didactics of Mathematics. Its main characteristics will be identified: the
phenomenon of formalism as a research question, models of a cognitive net and their
use for diagnosis, re-education and prevention as the main aims, introspection as a
research method. Finally, a method of implementing research results will be presented.
Keywords: methodology, formalism, model.

1. Introduction and Framework

In this article, we will identify and illustrate the main characteristics of research done
within the Prague Seminars of Didactics of Mathematics led by M. Hejný. The research
is focussed on the thinking processes and teaching/learning processes in mathematics.
The majority of the research is of a qualitative nature and falls into what Schoenfeld
(1991) calls basic research into cognition. “By basic in this context I mean work whose
fundamental thrust is to inquire into the nature of thinking and learning processes, and
which is not driven in any obvious way by applications.”

Our research consists of four main parts: modelling, diagnosis, re-education and
prevention. Here only the first part will be presented. The following text will be written
at three levels. Firstly, we will give a general description of the methodology. Secondly,
the ideas presented in the general description will be illustrated by our research
projects. Lastly, some references to the relevant work done by other contributors to the
Prague Seminars will be given, especially in those parts of the methodology where our
European Research in Mathematics Education I.II: Group 7 286

own research has not yet been developed. In the next section, we will present briefly the
two research projects with which the main points will be illustrated.

2. Research Projects

2.1 A Student’s Solution to a Word Problem

- see Stehlikova, 1996, 1995a, 1995c,d -

The initial goal of the research was to analyse a student’s ability to solve word
problems mainly from the point of view of formalism (see part 3). A group of five word
problems was given to 11-year old children in five schools in the Czech Republic and
three schools in England. The method of atomic analysis was chosen to get deeper
insights into the students solving processes. The method was developed at the
Bratislava Seminars in Mathematical Education and is described in PhD thesis
(Stehlikova 1995a) and in (Stehlikova 1995c) in detail. Atomic analysis is based on two
ideas, atomisation of the solving process and comparative analysis. It can be
characterised as a method whose aim is to investigate the student’s intellectual
processes, i.e. the sequence of written ideas and mental steps which caused this
sequence and the causality of this mechanism. The decisive role is played by cognitive
and personal phenomena by means of which the thinking process is described. The
analysis of a student’s written solution is done with respect to one or more phenomena
“atom by atom” (i.e. small meaningful parts of the solution). The elaboration of atomic
analysis from the point of view of methodology is one of the most important results of
the research (see Stehlikova 1995a).

2.2 Ability to Construct a Mental Representation of a Structure

- see Kubinova et al., 1997; Stehlikova, 1997 -

It is well known that some students are able to create their own pictures of structures
for different parts of mathematics. The development of mathematics thinking in these
cases is not based on the growth of the structure but mainly on re-structuring of their
European Research in Mathematics Education I.II: Group 7 287

mathematical knowledge. On the other hand, other students only accumulate isolated
pieces of knowledge. Such knowledge suffers from the “disease” of formalism (see part
3). In order to remove formal understanding, the mechanisms for the process of
structuring must be discovered. The research project is therefore focussed on a
student’s ability to discover and build a structure. The question is how does a structure
grow and how the ability to recognise certain classifying characteristics in a system and
to judge their importance, is developed. Phenomena describing the process of
structuring are looked for.

As a research tool a non-standard arithmetic structure A2 which is new to a student
but which can be built in parallel with a structure the student already knows (Z as the
ring without zero divisors) is being used. In this way we aimed to minimise the chance
that a student’s previous knowledge will influence the research results. Concepts in the
structure of A2 are of two types: expected, i.e. the same or very similar to those in Z, and
surprising, i.e. different from Z, which contribute to a student’s motivation and
penetration of the structure.

Structure A2=(A2,Å,Ä) consists of the set A2={1,2,...,99} of 99 natural numbers and
two binary operations z-addition Å and z-multiplication Ä defined as follows: xÅy =
R(x+y) and xÄy = R(x•y) where R:N®N we call a reduction mapping. The reduction
mapping can be simply presented on the set of all three digit numbers ABC and four
digit numbers ABCD as R(100A+10B+C)=A+(10B+C), R(1000A+100B+10C+D)
=(10A+B)+(10C+D). For example 73Å69=R(142)=1+42=43, 81Ä 90=R(R(7290))=

3. Research Questions

Research questions may originate from schools or from previous research as “the
second most significant products [of research] are stimuli to other researchers and
teachers to test our conjectures for themselves in their own context” (Mason, 1998). In
the Prague Seminars, the former is often the case. Central to its research is the
phenomenon of formalism which “... is the most serious ‘disease’ of a student’s
mathematical knowledge” (Hejny et al. 1990). Formal knowledge can be characterised
European Research in Mathematics Education I.II: Group 7 288

in brief as follows: (1) Individual pieces of mathematical knowledge are weakly
connected or even disconnected in a person’s mind. They are stored in the memory as
more or less isolated data. Connections among them are poor. A student is not able to
use his/her knowledge in a different context, for instance for characterising other pieces
of knowledge. (2) Mathematical knowledge is not connected with real life, i.e. it is not
“processed” through the student’s experience. For instance, a student is not able to find
a model of a mathematical concept in real life. (3) A student’s mathematical thinking is
oriented towards the question ‘How?’ rather than ‘Why?’. He/she tends to solve a
problem by applying an algorithm previously learnt.

We speak about formal knowledge when a student can give a definition, an
algorithm or a theorem but does not understand it, nor can they apply it in a different
context. For example, Peter is asked in a lesson on analytic geometry to decide if all
rotations in plane form a group or not. He knows the definition of group and all the
criteria a binary operation must satisfy but cannot apply it in the context of geometry.
His knowledge of group theory suffers from the disease of formalism. He needs more
experience with different models of groups. On the other hand, we do not speak about
formalism when a new piece of knowledge is presented to a student. It only exists in
isolation at first and only after gaining more experience is it accommodated in the
existing structure. It is only at a developmental stage then, not the manifestation of

Formalism can attack individual pieces of knowledge but it can “spread like a
disease”. It can penetrate into to the level of abilities and create a learning strategy in
which a student prefers learning by rote to learning with understanding. If formalism
attacks inner beliefs, a student is convinced that he/she is not able to learn mathematics
with understanding and that he/she can only master learning by heart. When presented
with a task, he/she does not attempt to gain insight but rather tries to use any algorithms
by chance.

From the application point of view, the main aim of our research is a diagnosis,
re-education and prevention of formalism. This leads to the necessity to explore the
anatomy of the process of gaining knowledge from the standpoint of a student’s
thinking processes and to create a model.
European Research in Mathematics Education I.II: Group 7 289

4. How to Create Models

In this part, we will illustrate the methodology of creating models through identifying
phenomena, trace and mechanisms. Let us characterise what we mean by these terms.
Even though it will be done separately for each term, the distinction among them may
be blurred and they can, and usually are, investigated at the same time.

4.1 Phenomena

The study of the thinking process begins by seeking those elements which link this
process (so far unknown) to our knowledge of a cognitive structure. These elements
will be called phenomena. A phenomenon answers the question, “What?“. What we
should concentrate on? What is important in a solving process? The process of
identifying and later describing phenomena is long and specific methods are needed.
“The need to identify and describe various cognitive structures in all phases of
construction suggests methods such as the clinical interview and prolonged
observation, that permit us to make inferences about the structures that underlie
behaviour“ (Noddings, 1990).

Illustration: In research 2.1, students written work from a standardised experiment
was taken as the basis and the method of atomic analysis was used for eliciting relevant
phenomena. For some examples of phenomena in research 2.1 see (Stehlikova
1995a,c). In 2.2, interviews with students were given preference. In current research,
hardly anyone would deny the merits of clinical interview (see Ginsburg, 1981, for a
detailed elaboration of this) even though it has disadvantages. Here we will describe in
more detail the research process in research 2.2 (it will be written in ich-form). The
research consists of two parts — the preparatory stage (steps 1 – 4) and the research
itself (steps 5 – 11).

1. I, as an experimenter, discovered A2 through a set of problems. This process was
recorded for future use.
2. I prepared a set of problems for the interviews with students related to my
experience with the structure. The equations were chosen to show a wide range of
European Research in Mathematics Education I.II: Group 7 290

possibilities, including equations with more than one solution, no solution and those
with the neutral element 99.
3. I conducted and recorded the first interview with Mary. First I showed Mary the set
A2 as the set of the first 99 natural numbers. Then I introduced reduction using
several concrete examples. After this I showed Mary the operation of z-addition and
z-multiplication, first generally and then on a concrete example.
4. I transcribed and analysed the interview. I listed several mistakes I had made when
conducting the interview.
5. I prepared a second set of problems on the basis of my experience with the first
experiment. Here it is:
(a) Reduction: r(100)=1, r(224)=26, r(1020)=30, r(1326)=39, r(2899)=r(127)=28,
Task: Choose several natural numbers and reduce them.
(b) Additive and multiplicative problems: 6Å60, 99Å35, ... 6Ä9, 4Ä48, ... Task: Think
about a word problem for some of these problems.

Note: This task was first used as a good motivator for the student. But it has become
clear that it is a good task for discovering a neutral element. Martina found it when
formulating these word problems. “We have 6 apples and we get 60 more. How many
apples do we have altogether?” ”We have 68 apples and we get 97 more.” (There comes
the surprise, sometimes it is better not to get any more.) “We have 35 apples and we get
99 more”. (Nothing changes.)

c) Simple equations: xÅ6=92, 61Åx=4,... 3Äx=45, 50Äx=5... More difficult
equations: 3ÄxÅ2=83, 5ÄxÅ10=5.

Note: This was the preparation for the first interview. In fact, students needed more
time to complete these tasks (with variations mentioned in the note below). Later they
were given other tasks such as divisibility, powers, square roots, sequences, quadratic
equations etc.

6. Interviews were conducted. Each interview lasted between 30 to 60 minutes. Each
student came as many times as he/she wanted (after school). They are all future
mathematics teachers.
European Research in Mathematics Education I.II: Group 7 291

Note: During the interviews with different students the need to alter the way of
presenting problems, to change their order etc. according to the situation became clear.
So finally, only the first interview was more or less the same but the next interviews
differed for different students according to their progress.

7. All interviews were transcribed word by word including all pauses, interruptions
8. Using my experience working in the structure A2 I prepared a set of “important
stages” (what was important for me, what caused problems). These stages were used
as a basis for the first analysis of interviews (i.e. the analysis was done with respect
to one ”important stage” at a time while all others were suppressed). The first list
was as follows: (a) Reduction — Does a student understand it? What mistakes does
he/she make? Does he/she understand the inverse process, i.e. that is for example
15=r(114)=r(213)=...? How long is the line of reductions? (b) Subtraction — How
to introduce and define subtraction? Is it possible to do it without the knowledge of
algebra? (c) Did a student find out that some equations have more solutions? How?
(d) Does a student use the same strategy of solving equations as the author? Does
he/she solve all equations in the same way or not? (e) Did a student find that there
exists a neutral element? How?
9. “Important stages” were determined for individual students. The first list was
10.On the basis of several lists of “important stages” the first list of phenomena was
made (for the lack of space it will not be presented here). By phenomenon we mean
something which plays an important role in the process of discovering the structure
of A2.
11.The second analysis from the point of view of the first list of phenomena was done.
All new interviews were analysed from the point of view of this list. This produced a
new list of phenomena and this is also being augmented as the research progresses.
(The names of phenomena need further clarification, S1 = natural or whole
numbers, i.e. the known structure (N or Z with addition and N or Z with
multiplication), S2 = the structure of A2 i.e. the new structure).
European Research in Mathematics Education I.II: Group 7 292

(I) The connection of S1 and S2 on the level of objects: (a) clarification of the basic
elements, e.g. with what certainty a student works inside A2, does he/she work with
numbers bigger than 99 or not; (b) elements which have a different character in S1 and
S2, e.g. a neutral element (z-zero); (c) elements which are not present in S1 or S2, e.g.
divisors of zero in S2. (II) The connection of S1 and S2 on the level of operations:
reduction, inverse reduction, z-addition, z-multiplication, z-division, z-subtraction.
(III) The connection of S1 and S2 on the level of strategies, e.g. the strategy of solving
linear equations. (IV) The transfer of experience between S1 and S2: (a) with a test and
adjustment, e.g. a student tries the strategy of solving quadratic equations as known
from S1 and if it does not work tries to modify it; (b) formal (without test), e.g. it does
not occur to a student that there could be something wrong, he/she just uses the method
he/she knows; (c) no transfer (a student seems not to use his/her experience from S1 in
the situation where it seems evident). (V) Autonomous work, e.g. a student asks his/her
own questions: Does there exist an equation with 9 roots? Does a quadratic equation
have two solutions? Does there exist a general rule for solving simple linear equations
which would enable me just to look at it and know the answer (e.g. when I know one
solution, I can quickly give all other solutions)?

We believe that phenomena are necessary for model creation, moreover they also
contribute to the specification of research and exemplification of research aims. In the
case of research 2.2, which is quite broad, it is our aim to focus more narrowly on one or
several phenomena which we hope will prove “profitable“.

4.2 Trace

A trace answers the question, “How?” It is the description of a time sequence of mental
steps (states). A trace can be either local (at most within 12 hours, but usually about 30
minutes) or global which provides long term information (at least a week but usually
half a year or more).

Illustration: In 2.2, a “local“ trace describes what happened during a period of one
interview which usually lasted 30-45 minutes. A “global“ trace is what we intend to do
in the future, to describe the change of the ability to build a structure in at least two
years. In addition, this ability will be investigated with different age groups. The
European Research in Mathematics Education I.II: Group 7 293

technology of tracing was elaborated also in research 2.1 (the technology of recording a
student’s solving process from the point of view of both what was written down and
‘inner speech’) (see Stehlikova 1995a) and in (Kratochvilova, 1997) in combinatorial

4.3 Mechanisms

While a trace describes what happened, a mechanism explains why it happened. It
answers the question “Why?”, i.e. it identifies causes of the items described in the trace.
If we know the mechanisms of a student’s solving style, we are able to predict how
he/she will solve a problem and/or explain why it was solved in the way it was.

Illustration: One of the universal mechanisms which was discovered within the
Bratislava Seminars and which serves as a vehicle in our research, is the mechanism of
the process of gaining knowledge (motivation, the stage of separate models, the stage of
a universal model, a piece of knowledge, its crystallisation, its automation) (see Hejny
et al., 1990 ). In 2.2, we observed the mechanism of a student’s attitude towards a
mistake. Mary, having made a mistake, always looked for it in the original solution. On
the contrary, Ann never did this. She preferred to look for other ways to solve the

4.4 Models

By means of the three preceding terms a model can be formed. By a model, we mean a
projection of a certain situation or process, which is hidden, to a context which
visualises this situation or process. The model has to be quite universal. The question is
how big a sample of concrete examples the model can describe and to what depth it
goes. These two tendencies often go against each other. Today’s trend is by case study.
Fewer examples but having greater depth are preferred (see e.g. Schoenfeld, 1991).

Illustration: In research 2.1, we made a model - scheme based on the basic
dichotomy of calculational and semantic levels of solving word problems (see
Stehlikova 1995c,d). Through this model we are able to follow the thinking processes
European Research in Mathematics Education I.II: Group 7 294

(traces) of a particular student, describe them by means of discovered phenomena and
identify his/her difficulties. The research 2.2 is currently at the level of discovering
phenomena and trace but in the long-term perspective, the goal is to create a model
which will diagnose the ability to build a structure. The model will not be a simple one
because the research is not focussed on a single concept but on the whole structure.

Other related projects: A model of grasping word problems can be found in (Hejny,
1995a). For a model of solving processes in word problems, mainly from the point of
view of strategies, see (Novotna et al. 1994).

5. Inner Research

In the previous paragraphs it was shown how models are based on information gained
through a student’s written or oral work. But there is another important way of getting
information and that is introspection as another characteristic of the methodology built
within the Prague Seminars. In our opinion, our approach corresponds with inner
research in Mason (1998) who claims that “inner research is about developing
sensibility” and “I need to be sensitive to the structure, importance and techniques of a
topic if I am to assist others to alter the structure of their attention”. It is distinctive to the
Prague Seminars that researchers (and teachers) go through the similar situations
(usually at a higher level) as their students to get a deeper insight into a student’s
thinking processes (see also Littler et al. 1998).

Illustration: In research 2.2, the experimenter was first introduced into the structure
of A2 in a similar way to the students and could therefore presuppose many of the
difficulties (obstacles) encountered by students in the process of their penetrating the
structure. Moreover, through introspection a ’new’ use of A2 was discovered, namely
how it can help produce a meaningful and natural introduction to the basic concepts of
group algebra. It brought to the surface the question whether it is possible at the
university level to teach the structure of algebra (or any other structure) so that the
knowledge is not afflicted by formalism.
European Research in Mathematics Education I.II: Group 7 295

6. Implementing Research Results

There are vast numbers of articles written on implementing research results into
practice. Here we would like to present two ways which are used within the Prague
Seminars. The first way corresponds to the current trend in mathematics education.
Practising teachers are encouraged to co-operate with researchers.

The second way will be illustrated in more detail. The following citation
summarises nicely our approach. “In order to maximise the effectiveness of research
effort in mathematics education it behoves us to reflect on what we go through when we
undertake research, and to try to offer and support practitioners in undertaking similar
or analogous personal reconstruction of research activities“ (Mason, 1998). The
Department of Mathematics and the Didactics of Mathematics of the Faculty of
Education, Charles University organised a whole series of workshops (called
“Iniciativa“) in which practising teachers took part. The aim was to introduce and let
them experience some research techniques rather than share the research results with

Our workshop focussed on the method of atomic analysis and its use for changing a
teacher’s attitude towards assessment of their students’ work. We created a similar
climate for teachers as we, the researchers, had when using atomic analysis. Rather than
tell the teachers the basic principles of the method and its advantages for their practice
we encouraged them to do their own experiments and to analyse the students’
outcomes. We wanted them to see for themselves what merits the method had and how
they could use it in their teaching. The feedback from this project shows that for the
majority of them, their sensitivity towards their students and especially their work,
increased (Stehlikova, 1995b). For similar projects see (Hejny, 1995b; Koman et al.
1995; Kubinova, 1995; Novotna, 1995).

7. Conclusion

We have identified and illustrated some distinctive features of the methodology
elaborated within the Prague Seminars. We have indicated directions in which our
European Research in Mathematics Education I.II: Group 7 296

research project will proceed. Several other models are being created using the same
methodology and it is our task to make an overall project analysis using the
comparative method.

8. References
Ginsburg, H. (1981): The Clinical Interview in Psychological Research on Mathematical Thinking:
Aims, Rationales, Techniques. For the Learning of Mathematics, 1 (3): 4-11.
Hejny, M. et al. (1990): Teoria vyucovania matematiky 2 (The Theory of Teaching Mathematics).
SPN Bratislava.
Hejny, M. (1995a): Zmocnovani se slovni úlohy. (Grasping a Word Problem.) Pedagogika XLV,
Hejny, M. (1995): Matematicke hry. (Mathematical Games.) See *.
Koman, M., Ticha, M. (1995): Jak ucit zaky tvorivemu pristupu k situacim s namety z praxe. (How to
Lead Students to A Creative Approach to Situations with Topics from Practice.) See *.
Kratochvilova, J. (1997): Thinking Processes Involved in Solving Combinatorial Problems. In M.
Hejny and J. Novotna (Eds.), Proceedings of ERCME 97. Prometheus Publishing House: 63-66.
Kubinova, M. (1995): Metodika experimentu. (Methodology of Experiments.) See *.
Kubinova, M., Stehlikova, N. (1997): Structures. In N. Stehlikova (Ed.), Didactics of Mathematics.
Proceedings of the Seminar of Doctoral Students. Rimov: 20-21.
Littler, G.H., Koman, M. (1998): Challenging Activities for Students and Teachers.
Mason, J. (1998): Researching from the Inside in Mathematics Education. In . Sierpinska and
J. Kilpatrick (Eds.), Mathematics Education as a Research Domain: A Search for Identity. An
ICMI Study. A, 357-377. Kluwer Academic Publishers.
Noddings, N. (1990): Constructivism in Mathematics Education. In Constructivist Views on the
Teaching and Learning of Mathematics. Journal for Research in Mathematics Education.
Monograph No 4: 7-18.
Novotna, J. (1995): Slovni rovnice a strategie jejich reseni. (Word Problems and Strategies of Their
Solutions.) See *.
Novotna, J., Kubinova, M. (1994): Strategie zakovskych reseni algebraicko-aritmetickych uloh.
(Strategies of Students’ Solutions to Algebraic-Arithmetical Problems.). In 17th World
Congress of Society for Science and Arts, Prague: page 81.
Schoenfeld, A. H. (1991): On Pure and Applied Research in Mathematics Education. Journal of
Mathematical Behaviour 10, 263-276.
European Research in Mathematics Education I.II: Group 7 297

Stehlikova, N. (1995): Analyza pisemneho resení slovni ulohy (zaku 5. tridy) (Analysis of a Student’s
Written Solution to a Word Problem). PhD thesis. Faculty of Education, Charles University,
Stehlikova, N. (1995): Analyza zakovskych pisemnych reseni. (Analysis of Students’ Written Work.)
See *.
Stehlikova, N. (1995): How Children Solved a Mathematical Word Problem: an Analysis. In ADUC
Bratislava: 33-54.
Stehlikova, N. (1995): The Solving Scheme as a Tool for Understanding Pupils’ Solutions. In
J. Novotna and M. Hejny (Eds.), Proceedings SEMT95, 117-120. Charles University, Prague.
Stehlikova, N. (1996): The Analysis of a Solution to the Word Problem. In M. Koman (Ed.),
Didactics, History, Fundamentals of Mathematics. Proceedings of the seminar on PhD studies
in the Czech Republic, 15-16. Kostelec nad Cernymi lesy.
Stehlikova, N. (1997): Phenomena of Structurization. In M. Hejny and J. Novotna (Eds.),
Proceedings of ERCME 97. 105-109. Prometheus Publishing House.

* (1995) Material for teachers in the project ’Iniciativa’ (described in part 6). Faculty of Education,
Charles University.

The contribution was financially aided by the grant GACR No. 406/97/P132
“Structuring mathematical knowledge“ and the grant GAUK No. 302/1998/A PP/PedF
“The phenomena of the mathematical part of the cognitive structure“. It was supported
by the Ministry of Education of the Czech Republic (Project PG 98374).
European Research in Mathematics Education I.II 298

Bagni, Giorgio . . . . . . . . . . . . . . . . . . . . 7-220
Batanero, Carmen . . . . . . . . . . . . . . . . . . 7-232
Bazzini, Luciana . . . . . . . . . . . . . . . . . . . 6-112
Bergsten, Christer . . . . . . . . . . . . . . . . . . 6-123
Boero, Paolo . . . . . . . . . . . . . . . . . . . . . 6-106
Bolea, Pilar . . . . . . . . . . . . . . . . . . . . . . 6-135
Bosch, Marianna . . . . . . . . . . . . . . . . . . . 6-135
Capponi, Bernard. . . . . . . . . . . . . . . . . . . 6-146
D’Amore, Bruno . . . . . . . . . . . . . . . . . . . 7-257
Ferrari, Pier Luigi . . . . . . . . . . . . . . . . . . 5-24
Gascón, Josep . . . . . . . . . . . . . . . . . . . . 6-135
Gelfman, Emanuila. . . . . . . . . . . . . . . . . . 5-38, 5-49
Geneves, B.. . . . . . . . . . . . . . . . . . . . . . 6-146
Godino, Juan Diaz . . . . . . . . . . . . . . . . . . 7-211, 7-232
González Marí, José Luis . . . . . . . . . . . . . . 7-245
Hejny, Milan . . . . . . . . . . . . . . . . . . . . . 7-211
Hershkovitz, Sara. . . . . . . . . . . . . . . . . . . 5-64
Hoyos, Veronica . . . . . . . . . . . . . . . . . . . 6-146
Iaderosa, Rosa . . . . . . . . . . . . . . . . . . . . 6-159
Kholodnaya, Marina . . . . . . . . . . . . . . . . . 5-38
Kubinova, Marie . . . . . . . . . . . . . . . . . . . 5-53
Littler, Graham . . . . . . . . . . . . . . . . . . . . 5-53
Lozinskaia, Raissa . . . . . . . . . . . . . . . . . . 5-49
Maier, Herman . . . . . . . . . . . . . . . . . . . . 7-211, 7-257
Malara, Nicolina . . . . . . . . . . . . . . . . . . . 6-159
Nesher, Pearla . . . . . . . . . . . . . . . . . . . . 5-64
Novotna, Jarmila . . . . . . . . . . . . . . . . . . . 5-53, 5-64
Oleinik, Tatyana . . . . . . . . . . . . . . . . . . . 5-75
European Research in Mathematics Education I.II 299

Poli, Paola . . . . . . . . . . . . . . . . . . . . . . 5-97
Reggiani, Maria . . . . . . . . . . . . . . . . . . . 6-172
Remezova, Galina . . . . . . . . . . . . . . . . . . 5-49
Rososhek,Sam . . . . . . . . . . . . . . . . . . . . 6-184
Rouchier, André . . . . . . . . . . . . . . . . . . . 7-275
Schwank, Inge . . . . . . . . . . . . . . . . . . . . 5-84
Shiu, Christine . . . . . . . . . . . . . . . . . . . . 7-211
Stehlíková, Nada . . . . . . . . . . . . . . . . . . . 7-285
Szeredi, Éva . . . . . . . . . . . . . . . . . . . . . 6-195
Török, Judit. . . . . . . . . . . . . . . . . . . . . . 6-195
Zan, Rosetta . . . . . . . . . . . . . . . . . . . . . 5-97
European Research in Mathematics Education I.II 300


algebra . . . . . . . . . . . . . . . . . . . . . . . 6-112, 6-146, 6-195, 7-220
algebra teaching. . . . . . . . . . . . . . . . . . . 6-159
algebraic structures . . . . . . . . . . . . . . . . . 6-184
algebraization . . . . . . . . . . . . . . . . . . . . 6-135
analysis . . . . . . . . . . . . . . . . . . . . . . . 7-257

beliefs . . . . . . . . . . . . . . . . . . . . . . . . 5-97

cognition . . . . . . . . . . . . . . . . . . . . . . 5-84
cognitive approach . . . . . . . . . . . . . . . . . 5-75
cognitive competence . . . . . . . . . . . . . . . . 5-38
comparison . . . . . . . . . . . . . . . . . . . . . 6-195
computer visualization . . . . . . . . . . . . . . . 6-184

didactical situations . . . . . . . . . . . . . . . . . 7-275
dynamic geometry . . . . . . . . . . . . . . . . . 6-146

EEG . . . . . . . . . . . . . . . . . . . . . . . . . 5-84
epistemology of mathematics . . . . . . . . . . . . 7-245
eye movement. . . . . . . . . . . . . . . . . . . . 5-84

formalism . . . . . . . . . . . . . . . . . . . . . . 7-285
European Research in Mathematics Education I.II 301

group . . . . . . . . . . . . . . . . . . . . . . . . 7-220

history . . . . . . . . . . . . . . . . . . . . . . . . 7-220

individualization of intellectual activity . . . . . . 5-38
institutionalization . . . . . . . . . . . . . . . . . 7-275
investigative methods . . . . . . . . . . . . . . . . 5-53

language. . . . . . . . . . . . . . . . . . . . . . . 5-24
learning syntactical . . . . . . . . . . . . . . . . . 6-159

mathematic organisation . . . . . . . . . . . . . . 6-135
mathematical thinking and learning . . . . . . . . 5-75
mathematics education . . . . . . . . . . . . . . . 7-232
meanings . . . . . . . . . . . . . . . . . . . . . . 6-172
meta-analysis . . . . . . . . . . . . . . . . . . . . 7-245
methodology . . . . . . . . . . . . . . . . . . . . 7-245, 7-285
microworld teaching . . . . . . . . . . . . . . . . 6-146
model . . . . . . . . . . . . . . . . . . . . . . . . 7-285
modelling . . . . . . . . . . . . . . . . . . . . . . 6-184
multiplicative compare problems . . . . . . . . . . 5-64

natural language. . . . . . . . . . . . . . . . . . . 6-112
European Research in Mathematics Education I.II 302

ontology . . . . . . . . . . . . . . . . . . . . . . . 7-232

performances . . . . . . . . . . . . . . . . . . . . 6-195
pragmatics. . . . . . . . . . . . . . . . . . . . . . 5-24
problem-solving. . . . . . . . . . . . . . . . . . . 5-24, 5-97
process of study . . . . . . . . . . . . . . . . . . . 6-135

regulation . . . . . . . . . . . . . . . . . . . . . . 7-275
rules . . . . . . . . . . . . . . . . . . . . . . . . . 6-172

school problems. . . . . . . . . . . . . . . . . . . 5-97
semiotics . . . . . . . . . . . . . . . . . . . . . . 7-232
structural aspects . . . . . . . . . . . . . . . . . . 6-159
symbols . . . . . . . . . . . . . . . . . . . . . . . 6-112, 6-172

teacher studies. . . . . . . . . . . . . . . . . . . . 5-75
textual eigenproductions . . . . . . . . . . . . . . 7-257
tools . . . . . . . . . . . . . . . . . . . . . . . . . 6-195

understanding . . . . . . . . . . . . . . . . . . . . 5-53

ways of information coding. . . . . . . . . . . . . 5-38